Preclinical relevance of probiotics in type 2 diabetes: A systematic review

Summary

Type 2 diabetes (T2DM) is among the most prevalent metabolic diseases in the world and may result in several long‐term complications. The crosstalk between gut microbiota and host metabolism is closely related to T2DM. Currently, fragmented data hamper defining the relationship between probiotics and T2DM. This systematic review aimed at investigating the effects of probiotics on T2DM in animal models. We systematically reviewed preclinical evidences using PubMed/MEDLINE and Scopus databases, recovering 24 original articles published until September 27th, 2019. This systematic review was performed according to PRISMA guidelines. We included experimental studies with animal models reporting the effects of probiotics on T2DM. Studies were sorted by characteristics of publications, animal models, performed analyses, probiotic used and interventions. Bias analysis and methodological quality assessments were examined through the SYRCLE's Risk of Bias tool. Probiotics improved T2DM in 96% of the studies. Most studies (96%) used Lactobacillus strains, and all of them led to improved glycaemia. All studies used rodents as models, and male animals were preferred over females. Results suggest that probiotics have a beneficial effect in T2DM animals and could be used as a supporting alternative in the disease treatment. Considering a detailed evaluation of the reporting and methodological quality, the current preclinical evidence is at high risk of bias. We hope that our critical analysis will be useful in mitigating the sources of bias in further studies.

1. INTRODUCTION

Diabetes mellitus is currently one of the most prevalent metabolic diseases in the world, affecting over 422 million people. ¹ Type 2 diabetes (T2DM) accounts for 90% of the cases, ¹ and it may be caused by impaired insulin secretion by the pancreas, decreased insulin sensitivity in target tissues or a combination of these elements. ² Consequently, there is a poor metabolic effect of insulin and chronic hyperglycaemia, ² which can result in microvascular complications, particularly in the heart, kidneys and nervous system in the long term. ³ Thus, its treatment is essential for preventing the development of comorbidities in patients with T2DM. ⁴ It has been reported that patients with T2DM may show altered gut microbial composition, ⁵ which suggests that modifying gut microbiota through probiotic treatment may be a way of regulating glucose metabolism. Although many researchers have suggested the hypoglycaemic potential of probiotics, the underlying mechanisms remain a matter of debate, as T2DM has a complex etiopathology. ⁶

Several microorganisms including bacteria, virus, fungi and archaea reside in the gut and mutually interact with the host. This relationship is influenced by both host's genetics and the intestinal environment, including diet and diseases. ⁷ Important enteric residents include the genera Lactobacillus, Bifidobacterium, Clostridium, Bacteroides, Faecalibacterium, Roseburia, Suterella, Akkermansia and Eubacterium. ⁷ For general human health, probiotic manufacturers recommend 1 capsule (10 billion CFU) per day of Lactobacillus rhamnosus GG. The crosstalk between microbiota and metabolic processes has to be accurately regulated in order to maintain physiological homeostasis. ⁸ It has been suggested that probiotics contribute to attenuate several elements of metabolic syndrome, in many cases related to T2DM. ⁸ These improvements may be due to inhibition of potential pathogens by the maintenance of a variety of mucosal immune cells, ⁷ decreased levels of pro‐inflammatory cytokines (ie tumour necrosis factor‐α and interleukin‐6), delayed glucose absorption in the gut and suppression of rate‐controlling enzymes of gluconeogenesis, such as glucose‐6‐phosphatase (G‐6‐Pase) and phosphoenolpyruvate carboxykinase (PEPCK). ⁴ , ⁶ Uncontrolled gluconeogenesis leads to hyperglycaemia in T2DM patients. Therefore, a suppression of G6Pase and PEPCK by probiotic treatment improves glycaemia by decreasing glucose release from the liver to the bloodstream. ⁴

Another mechanism involved with T2DM and its complications are overproduction of reactive oxygen species (ROS), which can impair insulin sensitivity and directly damage pancreatic β‐cells. ¹⁰ Recent studies propose that oral administration of multi‐strain probiotics enhances glucose tolerance, increases antioxidant activity and decreases chronic inflammation. ¹⁰ A better oxidative stress balance was also observed in diabetic animals treated with Lactobacillus strains, particularly in the liver. ⁴

Considering the extensive methodological variability in studies involving T2DM and probiotics, compiling all the current evidence is relevant so as to clarify the underlying mechanisms potentially linked to the relevance and applicability of probiotics in the treatment of T2DM. In this context, this systematic review was delimited as a tool to evaluate the variability between the studies available, allowing us to critically analyse the current evidence on the effects of probiotics in animal models of T2DM. This review also evaluates the methodological quality of the current evidence, indicating the main sources of bias associated with this current evidence.

2. METHODS

The systematic review adhered to the PRISMA guideline (Preferred Reporting Items for Systematic Reviews and Meta‐Analysis Protocols), ¹¹ including search strategy, selection criteria, data extraction and analysis.

2.1. Data sources and research records

To identify relevant articles, we searched the electronic databases PubMed/MEDLINE (https://www.ncbi.nlm.nih.gov/pubmed) and Scopus (https://www.scopus.com/home.uri). The studies considered eligible were identified until 27 September 2019. The keywords used in search filters were organized in three groups: (a) type 2 diabetes mellitus, (b) probiotics and (c) animal studies. Indexed studies were initially retrieved from search filters developed for PubMed by combining keywords and standardized MeSH descriptors (Medical Subject Headings, www.ncbi.nlm.nih.gov/mesh). In addition, the command TIAB (title and abstract) was applied to identify recently published records still in the indexing process. The logical operators ‘AND’ or ‘OR’ were used to combine all descriptors. To detect in vivo preclinical studies in PubMed, a standardized animal filter was applied. ¹² The search strategy applied to PubMed/MEDLINE was adapted for studies selection in Scopus, and the animal filter was provided by the search platform itself (Table S1). The references of eligible studies were checked for additional articles not identified by the electronic search (Figure 1).

FIGURE 1.

Flow diagram of the systematic review literature search results. Based on ‘Preferred Reporting Items for Systematic Reviews and Meta‐Analyses: The PRISMA Statement’. http://www.prisma‐statement.org. From: Moher D, Liberati A, Tetzlaff J, Altman DG, The PRISMA Group (2009) ¹¹

2.2. Records screening and eligibility

Studies had to conform with several criteria to be included in this systematic review, following the PICOS strategy (Table S2). The initial selection was based on title and abstract. After screening, duplicate studies and studies without experimental design were excluded. We considered only experimental studies performed in vivo, published in English and with full text available. Studies in vitro, reviews, comments, and notes performed in non‐diabetic animals or that aimed to assess effects of a medicine were excluded. We selected only studies that met all of the eligibility criteria as follows: 1) studies with type 2 diabetic animals and studies reporting the effects of probiotic use in T2DM.

The studies identified in this first screening were read in full and assessed for compliance with the established eligibility criteria. Requests to authors were made when the study was initially unavailable. An initial screening of search results was done by one reviewer to exclude clearly irrelevant records. The remaining records were screened by two independent reviewers, which identified potentially relevant records meeting the inclusion/exclusion criteria based on title and abstract analysis. Full papers were obtained from these records and were assessed for relevance by two independent reviewers. Any disagreements were resolved by consulting a third reviewer. The kappa test was performed to evaluate the agreement between two researchers, which indicated an almost perfect agreement (kappa = 0.917).

2.3. Data extraction and studies characteristics

Data extraction was based on methodological features, and the studies were synthesized admitting different descriptive levels as it follows: (a) publication characteristics (authors, year, country of origin); (b) characteristics of the animal models: species, strain, number of animals, sex, age and weight (Table S3); and (c) performed analyses, probiotic used and intervention, type of diabetes induction, and time of treatment. The primary outcomes were as follows: CFU = colony‐forming units; FBG = fasting blood glucose; and ppBG = postprandial blood glucose. The secondary outcomes were as follows: oxidative stress markers, serum insulin and body weight (Table S4). In the absence of available data within the study, authors were contacted via e‐mail to provide further information. Studies were initially grouped based on type/strain of probiotics used for treatment.

2.4. Risk of bias assessment

To assess the risk of bias in all studies included, the SYRCLE's Risk of Bias (RoB) tool specifically designed for animal studies was used. ¹³ The following methodological domains based on RoB were evaluated: Consider selection bias: ‘Was the allocation sequence adequately generated and applied?’, ‘Were the groups similar at baseline or were they adjusted for confounders in the analysis?’, ‘Was the allocation to the different groups adequately concealed?’; Consider performance bias: ‘Were the animals randomly housed during the experiment?’, ‘Were the caregivers and/or investigators blinded from knowledge regarding which intervention each animal received during the experiment?’; Consider detection bias: ‘Were animals selected at random for outcome assessment?’, ‘Was the outcome assessor blinded?’; Consider attrition bias: ‘Were incomplete outcome data adequately addressed?’; Consider reporting bias: ‘Are reports of the study free of selective outcome reporting?’; Consider other biases: ‘Was the study apparently free of other problems that could result in high risk of bias?’; and the overall study quality indicators: ‘Was it stated that the experiment was randomized at any level?’ and ‘Was it stated that the experiment was blinded at any level?’. The items in the RoB tool were scored with ‘yes’ (low risk of bias); ‘no’ (high risk of bias); or ‘unclear’ (indicating that the item was not reported, and therefore, the risk of bias was unknown). Reporting quality was evaluated by screening all manuscript sections (abstract to acknowledgements and funding). The analysis of each study and its relative risk of bias and year of publication were performed according to the methodological items reported in the SYRCLE's Risk of Bias (RoB) tool. ¹³

3. RESULTS

3.1. Characteristics of publications

The PRISMA diagram illustrates the process of studies selection (Figure 1). A total of 919 records were retrieved, with 250 from PubMed/MEDLINE and 669 from Scopus. Out of these, 683 remained after removing duplicates. Following assessment based on title and abstract, 98 records were considered to be suitable for potential inclusion, Inclusion criteria were met by 21 articles. The reference list of 21 selected articles was manually analysed, and 3 studies were added in accordance with the inclusion criteria, which brought us to 24 studies. Most studies originated from China (n = 15, 63%), followed by Japan (n = 2, 8%), Taiwan (n = 2, 8%) and India (n = 2, 8%). Other included studies were from South Korea (n = 1, 4%), Iran (n = 1, 4%) and Malaysia (n = 1, 4%).

3.2. Experimental animals and intervention characteristics

As shown in Table S3, rats (Rattus norvegicus) were the main animal model used (54%, n = 13) and 46% (n = 11) used mice (Mus musculus). Among the animal models used in the studies, 87.5% (n = 21) were males and 12.5% (n = 3) of the studies did not specify the animal gender. There were no studies reporting the use of female animal models. In regard to mice lineages, 33% of the studies used C57BL/6J (n = 8) and 4% of the studies used C57BL/KS/J and KK‐Ay mice (n = 1). Among rat lineages, most studies preferred Wistar rats (29%, n = 7), followed by Sprague‐Dawley (16%, n = 4) and GK rats (4%, n = 1). All animals were aged from 3 to 8 weeks. Mice weighted from 16 to 40 g, rats weighted from 120 to 230 g, and 10 studies did not report experimental animals' body weight. The number of experimental animals ranged from 18 to 105, and 14% of the studies (n = 3) did not provide this information.

The duration of probiotic treatment ranged from 3 to 18 weeks. Lactobacillus strains were investigated by 92% (n = 22) ⁴ , ⁶ , ¹⁰ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²¹ , ²² , ²³ , ²⁴ , ²⁵ , ²⁶ , ²⁷ of the studies, from which 21% (n = 5) were in combination with other probiotics from the genera: Bifidobacterium (n = 1), ²⁸ Pediococcus and Lactococcus (n = 1), ²⁹ Bacillus (n = 1), ³⁰ Rhizopus (n = 1) ³¹ and yeasts (n = 1). ³⁴ Other reported probiotics were Akkermansia muciniphila (n = 1) ³³ and Clostridium butyricum (n = 1). ³⁴ Oral administration of probiotics was used by 46% of the studies (n = 11) ⁴ , ⁶ , ¹⁴ , ¹⁵ , ¹⁶ , ¹⁷ , ¹⁸ , ¹⁹ , ²⁰ , ²⁴ , ²⁵ whereas 29% (n = 7) ²¹ , ²³ , ²⁶ , ²⁷ , ²⁸ , ³³ , ³⁴ used oral by gavage and the remaining 25% used probiotics in foods (n = 6). The foods used for probiotic administration were Shubat (n = 1), ³¹ Tempeh (n = 1), ³⁰ fermented paste Xenji™ (n = 1), ²⁹ fermented juice (n = 2) ¹⁹ , ²⁵ and fermented milk (n = 1). ¹⁰ The administered dose ranged from 10 to 10¹⁰ colony‐forming units (CFU) per mL regardless of the probiotic, although 3 studies ⁶ , ²⁵ , ²⁸ did not report the CFU per mL given as treatment for animals with T2DM. The time of probiotic treatment in the included studies ranged from 4 to 18 weeks.

Although it was not included as eligibility criteria, all studies included reported at least one of the following parameters of glycaemic control: fasting blood glucose (FBG), postprandial blood glucose (ppBG), glycated haemoglobin A1c (HbA1c) or oral glucose tolerance tests (oGTT).

4. MAIN FINDINGS

Preclinical studies demonstrated that probiotic intervention presents beneficial effects for T2DM treatment, as 96% (n = 23) of the studies reported lower glycaemia. A dose‐dependent benefit was reported in 13% (n = 3) of the studies. ⁴ , ¹⁴ , ¹⁶ The remaining 4% (n = 1) showed no difference in T2DM treatment. All studies that used Lactobacillus strains alone or associated with other microorganisms as probiotics (n = 22.92%) resulted in better T2DM parameters after treatment. Underlying mechanisms involved increased glucose tolerance (n = 16.67%), improved inflammatory status (n = 10.42%), increased antioxidant activity (n = 12.50%), delayed gastric absorption of glucose (n = 1.4%) and decreased gluconeogenesis (n = 3.13%). The effects of different probiotics on T2DM parameters in animal models are summarized in Table 1.

TABLE 1.

Summary of the effects of different probiotics on main parameters of T2DM in animal models

Probiotic	Effect	Measure outcomes
Lactobacillus sp (n = 17) 4 , 6 , 10 , 14 , 15 , 16 , 17 , 18 , 19 , 20 , 21 , 22 , 23 , 24 , 25 , 26 , 27	↓glycaemia (n = 17)	Fasting BG ppBG HbA1c oGTT
Bifidobacterium bifidum Lactobacillus casei (n = 1) 28	↓glycaemia ↑glucose tolerance (n = 1)	Fasting BG HbA1c oGTT
Lactobacillus sp (n = 11) 4 , 10 , 14 , 16 , 19 , 20 , 21 , 22 , 23 , 24 , 25	↑glucose tolerance (n = 11)	oGTT
Lactobacillus sp Leuconostoc mesenteroides Pediococcus acidilactici Pediococcus pentosaceus (n = 1) 29	↑glucose tolerance ↓antioxidant activity ↓cytokine production (n = 1)	oGTT SOD, CAT Fasting BG PCR SOD, CAT
Lactobacillus sp Bacillus subtilis (n = 1) 30	↓glucose tolerance ↓glycaemia (n = 1)	oGTT HbA1c Fasting BG
Lactobacillus plantarum Rhizopus oligosporus (n = 1) 31	↓glucose tolerance ↓glycaemia (n = 1)	oGTT HbA1c Fasting BG
Lactobacillus sp yeasts (Kluyveromyces marxianus, Pichia membranifaciens, Candida ethanolica, Issatchenkia orientalis) (n = 1) 32	↓glycaemia (n = 1)	Fasting BG HbA1c
Akkermansia muciniphila (n = 1) 33	No difference in glycaemia ↓cytokine production ↑antioxidant activity (n = 1)	Fasting BG Fasting serum levels of tumour necrosis factor‐α, interleukin‐6 SOD, GSH
Clostridium butyricum (n = 1) 34	↓glycaemia ↑glucose tolerance ↓cytokine production ↑antioxidant activity ↓expression of G6P genes (n = 1)	Fasting BG HbA1c oGTT Fasting serum levels of tumour necrosis factor‐α, interleukin‐6 SOD, CAT, GSH PCR
Lactobacillus sp (n = 7) 10 , 16 , 19 , 20 , 21 , 23 , 25	↓cytokine production ↑antioxidant activity (n = 7)	Fasting serum levels of tumour necrosis factor‐α, interleukin‐6 SOD, GSH
Lactobacillus sp (n = 2) 4 , 26	↓expression of gluconeogenic genes (n = 2)	PCR
Lactobacillus sp (n = 2) 24 , 27	↑antioxidant activity (n = 2)	SOD, CAT
Lactobacillus sp (n = 1) 6	↓absorption of glucose (n = 1)	oGTT

Data stratified by study are detailed in Table S2.

Abbreviations: BG, blood glucose; CAT, catalase; HbA1c, glycated haemoglobin A1c; oGTT, oral glucose tolerance test; ppBG, postprandial blood glucose; SOD, superoxide dismutase.

Regarding secondary outcomes, 16% (n = 4) of studies reported an increase in body weight, whereas 8% (n = 2) reported weight loss with probiotic treatment. Plasma insulin was increased in 16% (n = 4) of the included studies, and it was decreased in 16% (n = 4) of the studies. Food intake was decreased in 4% (n = 1) of the studies. Further details are found in Table S4.

4.1. Risk of bias methodological quality assessment

Quality assessment of the studies is summarized in Figure 2. Quality assessment reported at an individual level is shown in Figure 3. No studies fulfilled all methodological criteria analysed. In relation to selection bias, the sequence generation process was not fully reported in 92% (n = 22) studies (Q1). In terms of animals' characteristics, that is, their similarity to one another (Q2), 67% of the studies (n = 16) did not report this information clearly. Information about the allocation concealment (Q3) was not reported by 50% of the studies (n = 12). None of the articles reported on random housing and blinding of caregivers. Only 13% of the studies (n = 3) reported random animal housing (Q4). None of the studies reported blinding of personnel (Q5). Random selection for outcome assessment (Q6) was applied on 13% (n = 3) of the studies. None of the studies reported blinding of outcome assessment (Q7). Incomplete outcome data (Q8) were addressed in 71% (n = 17) of the studies. Selective reporting (Q9) was absent in 42% (n = 10) of the studies. An important source of bias was the lack of information about experimental models such as number of animals used, their sex and/or weight (Q10). Overall quality indicators, such as randomization (Q11) or blindness (Q12) at any level, were applied in 63% (n = 15) and 8% (n = 2) of the studies, respectively. None of the studies fulfilled all methodological criteria, and the mean quality score of all studies reviewed was 72.887 ± 3.623. Seven studies (29%) did not reach the mean score (Figure 3). Considering each criterion analysed individually, none of the studies reported information such as experimental blindness, a rational basis for the number of animals used and details of the sample size calculation. The analysis of the individual studies found no relation between risk of bias and year of publication (Figure 4).

FIGURE 2.

Results of the risk of bias and methodological quality indicators for all included studies in this systematic review that evaluated the effect of probiotic treatment in T2DM. The items in the Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE). Risk of Bias assessment were scored with ‘yes’ indicating low risk of bias, ‘no’ indicating high risk of bias, or ‘unclear’ indicating that the item was not reported, resulting in an unknown risk of bias. Consider selection bias: was the allocation sequence adequately generated and applied? Were the groups similar at baseline or were they adjusted for confounders in the analysis? Was the allocation to the different groups adequately concealed? Consider performance bias: were the animals randomly housed during the experiment? Were the caregivers and/or investigators blinded from knowledge which intervention each animal received during the experiment? Were animals selected at random for outcome assessment? Consider detection bias: was the outcome assessor blinded? Consider attrition bias: were incomplete outcome data adequately addressed? Consider reporting bias: are reports of the study free of selective outcome reporting? Consider other biases: was the study apparently free of other problems that could result in high risk of bias? Consider overall quality: was it stated that the experiment was randomized at any level? Was it stated that the experiment was blinded at any level? % represent the percentage of the studies that are fulfilled the requirements for ‘Low risk of bias’ and ‘Unclear risk of bias’ and ‘High risk of bias’

FIGURE 3.

Risk of bias summary shows studies' quality assessment at an individual level. (+) Low risk of bias. (−) High risk of bias. (?) Unclear risk of bias

FIGURE 4.

Analysis of methodological bias (reporting quality) stratified by domains for each study included in the review. Dotted line indicates the mean quality score (%)

5. DISCUSSION

Preclinical evidence of the relevance of probiotics in T2DM is of great relevance considering the high prevalence of this disease and the possibility of a new alternative therapy to improve life quality. In addition, as probiotics are abundantly available and affordable, they may represent an interesting supporting alternative for T2DM treatment. Despite the great heterogeneity of the studies included in this review, our findings suggest beneficial effects of probiotic in T2DM, mainly related to increased glucose tolerance, delayed glucose absorption and improved systemic inflammation, which are related to increased SCFA (short‐chain fatty acid) production. Other effects are related to increased antioxidant activity and decreased hepatic gluconeogenesis.

Regarding the origin of the studies included in this systematic review, all of them were developed in Asian countries, most of them from China. These data were already expected since as the use of probiotics for fermentation is a widespread tradition with a marked socio‐economic relevance in Asia. ⁴⁴ In China, it has been documented by several historians as far back as 4000 BC. ⁴⁴ Indeed, a typical diet of the Eastern World consists of boiled rice with many side dishes containing fermented foods, such as kimchi, soy sauce and miso. ⁴⁴ The use of fermented foods in Asian countries has been related to the relevance of probiotics in alleviating lactose intolerance for Asian people compared to lactose‐tolerant European people. Furthermore, the addition of Lactobacillus strains in foods is related to improved absorption of vitamins B2, B11 and B12, which is particularly relevant in developing Asian countries and where there is an ageing population. ⁴⁵

Besides the origin of the included studies, our results show that all experimental animals were murine models. This could be due to their low cost, and easy handling and accessibility, allowing researchers to use a relatively large number of animals for their experiments, thereby generating a greater degree of evidence and reliability in the results. In addition, none of the studies reported the use of female animals. Single sex studies of males still predominate in the biological literature, and neglect of females is widespread in many disciplines. The preference for male animal models could be due to the fact that female murine models need more anti‐inflammatory bacteria to maintain intestinal health, ³⁰ which could increase the amount of probiotics required or increase the time of probiotic treatment. These sex‐related differences may underlie dissimilarities in the epidemiology, aetiology, and prognosis of diseases. Given these fundamental differences, it seems essential to investigate probiotic treatment in both males and females. In addition, animals in weaning age were more frequently used compared to older ones, which may be due to better adaptation to treatments. ⁴⁶

Regarding the main results of this review, our findings suggest that SCFA production is responsible for several mechanisms that are reported to reduce blood glucose levels and improve overall health status. ³² SCFAs, such as butyrate, propionate and acetate, are metabolites produced by the gut microbiota as a result of microbial fermentation of oligo‐ or polysaccharides. SCFAs interact with host metabolism improving glucose tolerance due to increased glucagon‐like peptide‐1 (GLP‐1) levels and peptide YY (PYY). ³⁵ This is in accordance with our results as 67% (n = 16) of the studies showed improved glucose tolerance after probiotic treatment. In addition, enhanced GLP‐1 and PYY lead to increased satiety, ²⁶ , ³⁶ although decreased food intake was reported in 5% (n = 1) of the studies. ¹⁴ Furthermore, the increased secretion of GLP‐1 induced by SCFAs leads to delayed absorption of glucose, which results in better glycaemia. ³² , ³⁷ This is consistent with the results reported in the included studies, as 4% showed slower glucose absorption and lower postprandial blood glucose levels. After a meal, GLP‐1 is secreted and acts attenuating gastric emptying, modulating insulin production and reducing glucagon secretion. ³⁸ Therefore, enhanced GLP‐1 secretion due to probiotic treatment appears to benefit individuals with T2DM. ³⁸

In addition to these beneficial effects, the SCFAs produced by the gut microbiota also boost the innate immune response against pathogens, ⁹ inducing the activation, migration, proliferation, differentiation and maintenance of a variety of immune cells, as shown by 8% of the studies recovered in this review (n = 2). The SCFA butyrate induces regulatory T cells through the activation of G protein‐coupled receptors. Another SCFA, propionate, increases colonic regulatory T cells numbers by signalling via Ffar2. SCFAs also exert a direct bacteriostatic effect towards pathogens through the production of bacteriocins. ³⁹ Moreover, microbial groups produce vitamins such as B2, B11 and B12, ⁴⁰ which have been reported as deficient in T2DM individuals, ³⁸ thus improving nutrient availability and overall health. ⁴¹ Another mechanism related to SCFAs production improving T2DM management is a decreased inflammatory process. The onset and progression of metabolic diseases are closely related to upregulation of inflammatory cytokines, which are mainly induced by the binding of lipopolysaccharides (LPS) and TNF‐α to the surface of innate immune cells. ⁴¹ , ⁴² , ⁴³ SCFAs decrease cytokine production by preventing its gene expression, ⁴² hence alleviating the chronic inflammation of T2DM, ²³ as reported in our results. ¹⁰ , ¹⁶ , ¹⁹ , ²⁰ , ²¹ , ²³ , ²⁴ , ²⁵ , ²⁷ , ²⁹ , ³³ , ³⁴

In addition to increased inflammation, ⁴¹ T2DM patients often show high levels of oxidative stress indicators as a result of increased generation of free radicals, which may impair metabolic condition in T2DM. ³¹ Although the exact mechanisms remain unclear, attenuation of oxidative stress through probiotic treatment may also indirectly affect insulin level and glucose homoeostasis, ⁵ , ²⁷ as reported by 50% of the studies included in this review that showed increased antioxidant activity, mainly through increased SOD and CAT activities. ¹⁰ , ¹⁶ , ¹⁹ , ²⁰ , ²¹ , ²³ , ²⁴ , ²⁵ , ²⁷ , ²⁹ , ³³ , ³⁴

Finally, glucose homeostasis is directly influenced by gluconeogenesis, as an uncontrolled synthesis of hepatic glycogen may result in hyperglycaemia. ⁴ Rate‐controlling enzymes of gluconeogenesis PEPCK and G‐6‐Pase were suppressed after probiotic use, ⁴ which is consistent with the simultaneous increase in hepatic glycogen. ⁴ This is in accordance with our findings as 13% of the studies reported a downregulation of gluconeogenic genes and reduced blood glucose levels as a consequence. The secondary results of this review, such as plasma insulin and body weight, varied widely. We suggest that the specific rodent model (age of the animals and the mode by which metabolic disorders are induced), the treatment duration and type of probiotic selected influenced the outcomes of the included studies. ⁴³ Both young and adult animals were used as experimental animals, even though animals have different nutritional requirements throughout their lives. ⁴³ Furthermore, the treatment duration varied widely among studies (4‐18 weeks), and some metabolic changes may take several weeks to emerge, ⁴³ which could influence the results. In addition, distinct animal species may respond differently to the administered probiotics, as each organism has its own microbial population. ⁴³

Our main findings were summarized in Figure 5, which demonstrates the effects of probiotics in different organs. In the liver, decreased gluconeogenesis due to a decrease in PEPCK and G‐6‐Pase activity was reported. In the intestine, we observed an increased microorganism's population, leading to increased vitamin production. We also observed increased SCFAs production, resulting in an attenuated glucose absorption, increased glucose tolerance, increased satiety, decreased secretion of inflammatory cytokines and a boost in the immune system. Figure 5 also shows results of an increased antioxidant capacity, specially induced by upregulation of catalase and superoxide dismutase activities triggered by probiotics.

FIGURE 5.

Summary of the main findings on the effects of probiotic use in animals with T2DM. CAT, catalase; G‐6‐Pase, glucose 6‐phosphatase; GLP‐1, glucagon‐like peptide‐1; PEPCK, phosphoenolpyruvate carboxykinase; PYY, peptide YY; ROS, reactive oxygen species; SCFAs, short‐chain fatty acids; SOD, superoxide dismutase

Methodological consistency is crucial for evidence reliability in clinical and preclinical studies. ⁴⁷ Blindness of personnel and blinding of outcome assessment was not reported in the included studies, which impairs methodological quality in these studies. Other relevant sources of bias are under‐reporting randomization methods in the selected studies, data about experimental models such as age/weight/sex, number of animals used and the dose of probiotics used. Further, details in regards of group similarity at baseline were also under‐reported. Neglecting these methodological aspects lead to bias and compromise the quality of evidence. ⁴⁷ There are a number of guidelines on experimental design available in order to improve research quality and reliability, such as Approach Collaborative for Meta‐Analysis and Review of Animal Data from Experimental Studies (CAMARADES; www.camarades.info) and the Systematic Review Centre for Laboratory Animal Experimentation (SYRCLE; www.SYRCLE.nl). Following these guidelines improves research transparency and reproducibility, resulting in studies with a low risk of bias. In our review, we did not observe a direct relationship between the high risk of methodological bias and the year of publication of the studies, therefore there has been a systematic reproduction of methodological errors over the years. These findings show the need to enhance methodological quality of research in this area in order to reduce biases. We expect that this review can be used as a guide to improve reporting for future research regarding probiotic treatment in T2DM.

6. LIMITATIONS

Systematic reviews are considered as high‐level studies that allow for the individual evaluation of studies in a blind manner, using specific tools. ¹³ Such features result in an inclusive and reliable approach that provides a broad understanding of the included studies. Here, methodological discrepancies between the studies reviewed became clear considering the wide variety in animal models' characteristics such as age, weight and total number of animals. Additionally, the variability in probiotic dose is a possible caveat that may hinder conclusions that are more comprehensive. Our risk of bias and methodological quality assessment showed that many studies fail in reporting their complete methodology, resulting in a high risk of bias. We did not find a direct relationship between the high risk of methodological bias and the year of publication of the studies. Thus, it is likely that a continuous reproduction of methodological limitations might have occurred over the years, since the quality of the reports has not improved. For these reasons, we expect that this review can help improve reporting for future research on probiotic supplementation in T2DM.

7. CONCLUSION

There are multiple studies suggesting that probiotic treatment has an effect in reducing blood glucose levels and improving different metabolic parameters that affect glycaemia. Some potential mechanisms underlying these effects are improved glucose tolerance, better inflammatory status, decreased oxidative stress, delayed gastric absorption of glucose and decreased hepatic glycogen synthesis. However, current evidence is at a high risk of bias. Although improving research methods is needed for future studies, probiotic treatment represents a potential supporting alternative for T2DM management.

CONFLICT OF INTEREST

The authors declare no conflict of interest.

Supporting information

Table S1‐S4

Marques AM, Sarandy MM, Novaes RD, Gonçalves RV, Freitas MB. Preclinical relevance of probiotics in type 2 diabetes: A systematic review. Int. J. Exp. Path.. 2020;101:68–79. 10.1111/iep.12359