Can dietary fructans lower serum glucose?

Authors


Shanthi Johnson, Faculty of Kinesiology and Health Studies, University of Regina, 3737 Wascana Parkway, Regina, Saskatchewan, S4S 0A2, Canada.
Tel.: +1 306 337 3180
Fax: +1 306 585 4854
Email: shanthi.johnson@uregina.ca

Abstract

Background:  Convincing evidence indicates that the consumption of inulin-type fructans, inulin, and oligofructose has beneficial effects on blood glucose changes in animal models, although data in humans have been considered equivocal. As such, a systematic review of available literature on humans was conducted to evaluate the effectiveness of dietary inulin-type fructans on serum glucose.

Methods:  Thirteen eligible randomized controlled trials (RCT), published from 1984 to 2009, were identified using a comprehensive search strategy involving the PubMed, Medline, and Cochrane Library databases. Exclusion criteria, such as the absence of a control group, lack of information on the quantity of inulin-type fructans used, and lack of glucose values at outcome, were established.

Results:  Upon review, only four of the 13 trials (31%) showed a decrease in serum glucose concentration and only one of these was statistically significant. The remaining nine trials showed no significant changes in serum glucose concentration.

Conclusion:  Based on the present systematic review, it does not appear that inulin-type fructans have a significant lowering effect on serum glucose in humans. More RCT are needed to determine whether inulin-type fructans, inulin, and oligofructose have beneficial effects on blood glucose in humans.

Introduction

The worldwide prevalence of diabetes is projected to be over 285 million individuals1. The regulation of serum glucose concentrations in individuals is important for the prevention and management of the disease. Diabetes is diagnosed when serum glucose concentrations are >7.00 mmol/L, whereas normal fasting serum glucose is considered to be <5.55 mmol/L.2

An individual’s overall diet may have significant lowering effects on postprandial and fasting glucose concentrations.3–5 Current research indicates that soluble fiber in particular may have a lowering effect on serum glucose by reducing the postprandial glucose response.6 Furthermore, an association between dietary fiber and a reduced risk for developing chronic diseases, such as cardiovascular disease (CVD) and diabetes, has been shown.7 Given the benefits of fiber, current guidelines recommend that Canadians consume 26–35 g fiber daily; however, most consume between 4.5 and 11 g/day, far below the recommended values.8

Inulin is a type of fiber that has been shown to improve insulin concentrations in people with diabetes.9 Inulin belongs to the class of carbohydrates known as fructans10 and is a natural food ingredient found in over 36 000 flowering plants11, such as wheat, onion, banana, garlic, Jerusalem artichoke, and chicory root.12 Oligofructose, also known as fructo-oligosaccharide, belongs to the inulin family because it is a processed form of inulin that produces either short- or long-chain fructans.10 The different chain lengths are achieved through hydrolysis or physical separation techniques.13 Although inulin-type fructans are carbohydrates, their unique β2→1 linked glycosidic bonds cause them to be indigestible by human intestinal enzymes. These glycosidic bonds produce the reduced caloric value and dietary fiber effects of inulin-type fructans.12 Most inulin in the commercial world is synthesized from sucrose or extracted from the chicory root and Jerusalem artichoke.10 An average North American typically consumes 1–4 g inulin per day, mostly in the form of wheat, onion, garlic, and banana.14

Studies evaluating the effects of fructans on glucose concentration in humans have shown mixed results, although in animal studies the effects have been consistently positive. Specifically, a group of researchers conducted a series of studies on healthy beagles in which inulin was mixed with sugar beet fiber and added to the dogs’ diet. Results indicated that high concentrations of inulin may play a role in lowering blood glucose, insulin, and lipid levels and be useful in diets for dogs with diabetes.15,16 In another study observing the effects of life-long inulin supplementation on healthy rats, there was no change in fasting glycemia values at any point during the study in either the inulin-treated or control groups.17 Another comprehensive study in lean and obese rats assessed the effects of an inulin, high protein, or a combined fiber/protein (CB) diet on glucose tolerance and glucagon-like peptide (GLP)-1. Lean rats in the CB group had significantly higher GLP-1 levels than rats in the control and inulin groups.18 These results suggest that, in animals, inulin has glucose-lowering effects. However, the findings in humans have been equivocal. As such, the aim of the present systematic review of the available literature was to assess the effectiveness of inulin-type fructans on fasting serum glucose concentrations in humans.

Methods

For the systematic review, a comprehensive search strategy was established as shown in Fig. 1. A computerized search of all relevant randomized controlled trials (RCTs) published from 1984 to 2009 was performed in three databases: Pubmed, Medline, and the Cochrane Library. RCTs were used because these types of trials use the strongest type of research design compared with quasi-experimental and observational studies.19 The following keywords were used to identify primary articles: fructan, inulin, oligofructose, fructo-oligosaccharide, glucose, diabetes, blood glucose, and blood sugar. In addition, references listed in the articles retrieved were searched manually. Based on this search, 76 articles were identified as potentially suitable for inclusion, as shown in Fig. 1. All citations were exported to the reference software EndNote X for Windows (Thomson Reuters EndNote, Philadelphia, PA USA) for reference management. All available articles were reviewed for relevance. The Abstract, Introduction, Methods, and Results sections of each study were screened using the following inclusion criteria: (i) human study; (ii) specific types and amounts of inulin-type fructan are included; (iii) the study is an RCT; and (iv) measures of glucose concentration are a primary study outcome. Only articles written in English were included in the present review. The 76 articles were reviewed for inclusion by two individuals (N.B. and K.M.), independent of each other. After the initial independent review, one reviewer (N.B.) had selected 12 of 76 articles that fit the inclusion criteria, whereas the second reviewer (K.M.) had selected nine of 76 articles. The suitability of the three articles that were not selected by the second reviewer were discussed and the reviewers were able to come to the agreement that the three articles did indeed fit the inclusion criteria and should be included in the final set of articles for review. Thus, the final level of agreement reached between the two reviewers was 100% and 12 studies were included. One of the selected articles included two RCTs, resulting in a total of 13 trials within the 12 articles included in the present systematic review.

Figure 1.

 Flow chart summarizing the search process.

Results

Selected characteristics of the 13 trials are listed in Tables 1 and 2. Table 1 lists the characteristics of the trials in which serum glucose concentration decreased, whereas Table 2 list the characteristics of the trials in which there was no decrease in serum glucose concentration. Specific changes regarding serum glucose and lipid concentrations in all studies are given in Table 3. Overall, the 13 trials selected represented a total of 252 subjects. Six trials focused on men20–24, whereas the remaining seven trials included both sexes25–31. Six studies involved healthy participants,20–22,25,26 whereas the remaining seven trials were performed on individuals with hypercholesterolemia or hypertriglyceridemia,23,27,28 Type 2 diabetes,29–31 or non-alcoholic fatty liver disease.24

Table 1.   Trials in which decreases in fasting serum glucose levels were reported
ReferenceSubjectsFructanDose (g/day)Study designDuration (weeks)Method of fructan deliveryDecreases observed in:
GlucoseInsulin
  1. RCT, randomized controlled trial; normo-LP, normal lipoprotein levels; DB, double blind; NASH, non-alcoholic steatohepatitis; CO, crossover; FOS, fructo-oligosaccharide; parallel, parallel research design.

  2. Bolded entries indicate that diabetes was present in the study population.

  3. *< 0.05.

Yamashita et al.298 M/10 F; Type 2 diabeticsFOS8.0RCT, DB, parallel2Coffee drink and canned coffee jellyYes*Not studied
Jackson et al.2754 M&F; mod-LPInulin1000RCT, DB, parallel8Powder added to drinks and foodYesAt midpoint*
van Dokkum et al.2212 M; normo-LPInulin15.0RCT, DB CO3Orange juiceYesNot studied
Daubiouiul et al.247 M; NASHFOS16.0RCT, DB CO8Powder added to drinks and foodYesYes
Table 2.   Trials in which fasting serum glucose levels did not decrease
ReferenceSubjectsFructanDose (g/day)Study designDuration (weeks)Method of fructan deliveryDecreases observed in:
GlucoseInsulin
  1. RCT, randomized controlled trial; normo-LP, normal lipoprotein levels; hyper-LP, hypercholesterolemia or hypertriglyceridemia; DB, double blind; NASH, non-alcoholic steatohepatitis; CO, crossover; FOS, fructo-oligosaccharide; parallel, parallel research design.

  2. Bolded entries indicate that diabetes was present in the study population.

Luo et al.2012 M; normo-LPFOS20.0RCT, DB CO4.00100 g biscuitsNoNot studied
Schaafsma et al.2130 M; normo-LPFOS9.4RCT, DB, CO7.00YogurtNoNot studied
Alles et al.3020 M&F; Type 2 diabetesFOS15.0RCT, DB CO3.00PowderNoNot studied
van Dokkum et al.2212 M; normo-LPFOS15.0RCT, DB, CO3.00Orange juiceNoNot studied
Causey et al.2312 M; hyper-LPInulin20.0RCT, DB, CO3.00Ice creamNoYes
Luo et al.316 M/4 F; Type 2 diabetesFOS20.0RCT, DB, CO4.00Not shownNoNot studied
Letexier et al.254 M/4 F; normo-LPInulin (high performance)10.0RCT, DB, CO3.00Not shownNoNot studied
Giacco et al.2820 M/10 F; hyper-LPFOS10.6RCT, DB, CO24.00Powder mixed with tea/coffeeNoNot shown
Forcheron et al.2617 M&F; normo-LPInulin10.0RCT, DB, parallel24.00Powder in waterNoNo
Table 3.   Changes in serum lipids and glucose after the intervention period
ReferenceTotal cholesterolHDLLDLTriglyceridesGlucoseInsulin
ControlExplControlExplControlExplControlExplControlExplControlExpl
  1. HDL, high-density lipoprotein; LDL, low-density lipoprotein; Expl, experimental group; N/A, not available.

  2. Where appropriate, data are given as the mean (±SD) changes in mmol/L, unless noted otherwise.

  3. Significant changes are bolded.

Luo et al.203.91 ± 0.173.96 ± 0.221.05 ± 0.060.97 ± 0.05NANA0.72 ± 0.050.83 ± 0.164.86 ± 0.164.94 ± 0.1053 ± 462 ± 14
Alles et al.306.01 ± 1.186.21 ± 1.291.09 ± 0.251.12 ± 0.232.44 ± 1.043.99 ± 0.912.44 ± 0.792.56 ± 1.228.59 ± 2.668.61 ± 2.61NANA
Forcheron et al.263.73 ± 0.164.14 ± 0.161.13 ± 0.101.47 ± 0.112.31 ± 0.152.33 ± 0.200.64 ± 0.110.77 ± 0.143.81 ± 0.204.03 ± 0.106.80 ± 1.10 pmol/L6.80 ± 1.10 pmol/L
Luo et al.315.15 ± 0.245.13 ± 0.271.01 ± 0.061.02 ± 0.083.85 ± 0.203.85 ± 0.231.42 ± 0.121.33 ± 0.168.88 ± 0.518.89 ± 0.53118 ± 24 pmol/L115 ± 19 pmol/L
Schaafsma et al.215.175.411.231.213.523.331.571.525.375.47N/AN/A
van Dokkum et al.224.56 ± 0.784.51 ± 0.561.14 ± 0.221.16 ± 0.222.82 ± 0.512.81 ± 0.501.40 ± 0.681.31 ± 0.58NANANANA
Letexier et al.254.12 ± 0.324.35 ± 0.30Not shownNot shownNot shownNot shown0.92 ± 0.100.77 ± 0.084.62 ± 0.074.68 ± 0.147.90 ± 0.60 mU/L8.90 ± 1.40 mU/L
Yamashita et al.29245 ± 40 mg/dL223 ± 27 mg/dLNot shown50 ± 14 mg/dL168 ± 38 mg/dL147 ± 32 mg/dLNot shown146 ± 77 mg/dL198 ± 51 mg/dL182 ± 45 mg/dLNot shownNot shown
Jackson et al.276.46 ± 0.915.90 ± 0.971.31 ± 0.391.31 ± 0.334.43 ± 1.084.00 ± 0.851.59 ± 0.581.29 ± 0.354.99 ± 0.494.84 ± 0.5144.90 ± 27.50 pmol/L37.50 ± 18.30 pmol/L
Causey et al.23228 ± 40 mg/dL220 ± 32 mg/dL36.83 ± 9.95 mg/dL35.50 ± 7.81 mg/dL150 ± 32 mg/dL147 ± 27 mg/dL283 ± 194 mg/dL243 ± 162 mg/dLNot shownNot shown97.50 ± 58.88 mg/dL128 ± 55 mg/dL
Giacco et al.286.44 ± 0.786.47 ± 0.701.32 ± 0.391.29 ± 0.344.56 ± 0.784.58 ± 0.671.56 ± 0.531.53 ± 0.715.38 ± 0.835.44 ± 1.0050.20 ± 11.00 pmol/L51.70 ± 15.10 pmol/L
Daubioiul et al.24184 ± 13 mg/dL186 ± 12 mg/dL44.90 ± 5.40 mg/dL45.90 ± 5.30 mg/dL125 ± 13 mg/dL121 ± 10 mg/dL111 ± 18 mg/dL99 ± 12 mg/dL128 ± 17 mg/dL110 ± 7 mg/dL15.40 ± 3.10 mg/dL14.90 ± 4.00 mg/dL

As per the inclusion criteria established for the present systematic review, all 13 studies were designed as an RCT; however, three different types of RCT were used. Nine of the 13 trials used a crossover design,20–25,28,30,31 three used a parallel design,26,27,29 and oneused a sequential design.22 The test substance was oligofructose in eight trials20–22,24,28,30,31 and inulin of different chain lengths in four trials.23,25–27 One of the trials used a combination of both oligofructose and inulin.22 The average amount of active substance consumed was approximately 13.8 g/day. This can be broken down further to 13.2 g/day for oligofructose (range 5–20 g) and 12 g/day for inulin (range 5–20 g). In one trial,21 inulin-type fructans were consumed together with probiotic microorganisms. In the remaining trials, the test substance was administered either as a sweetener or powder consumed alone, added to drinks, added as an ingredient of biscuits,20 or added to orange juice.22 The trials varied in length, ranging from 14 to 168 days, with a mean duration of approximately 41 days.

Changes in fasting serum glucose and insulin levels

For the 13 trials, serum glucose concentration decreased in the group containing inulin-type fructans compared with the corresponding control group in four (31%) of the comparisons. Of these four comparisons, only one showed a significant change between treatment groups (see Table 1). The fructan consumption in these four trials varied between 8 and 16 g/day. The trial that showed the only significant decreases in glucose concentration used the least amount of fructans of 8 g/day.29 The duration of the trials that showed lowering effects of serum glucose ranged from 14 to 56 days. Although results from several other trials did not show significant decreases in serum glucose concentration, they did indicate trends towards improved health. One trial included in the analysis23 found that although there was a decrease in glucose concentration after inulin consumption, a trend towards increased insulin levels was noted and glucagon were significantly elevated. A subsequent trial included in the analysis26 also reported trends towards decreased fasting glucose concentrations among those receiving fructans. Another trial showed significantly reduced total body glucose disposal and hepatic glucose production.31 With regard to insulin, one trial27 showed a significant decrease in insulin levels at the midpoint of the trial. However, at final follow-up measurements, there was no significant difference from baseline measurements.

Discussion

To the authors’ knowledge, this is the first systematic review to focus solely on the effects of inulin-type fructans on fasting serum glucose concentrations in the human body. Systematic reviews hold many advantages over a traditional narrative review. Narrative reviews often do not include the reviewers’ search strategies and the quality of the studies is not considered. In contrast, systematic reviews involve an exhaustive search for primary studies based on a focused research question.32 In addition, the use of systematic reviews provides the reader with a high level of evidence when knowledge is sought regarding a specific topic.33

The purpose of the present systematic review was to determine the association between consumption of inulin-type fructans and fasting serum glucose concentrations. Based on the analysis of the studies included, it does not appear that consumption of inulin-type fructans has significant serum glucose concentration-lowering effects. This result is somewhat surprising because previous researchers have lauded the use of inulin and oligofructose as a fiber ingredients and sweetening agents.34 In addition, research has shown that diets rich in dietary fiber improve glycemic control and aid in the management of diabetes.35,36 The potential benefits of inulin-type fructans as a therapy for the management of diabetes and related glucose concentration requires further confirmation.

The results from the systematic review do not suggest that inulin-type fructans have no glucose-related health benefits or health benefits in general after consumption. The prebiotic, bifidogenic nature of inulin-type fructans has many non-glucose health-related benefits, such as reducing ammonia in the blood10 and greatly reducing constipation.37 Research has also indicated that inulin-type fructans may have serum lipid-lowering effects in humans, particularly those with hypercholesterolemia.38 Lipid levels are of importance because high serum lipid levels are often a complication in people with diabetes.39 The lowering of lipids occurs because inulin-type fructans are not hydrolysed by enzymes in the small intestine of humans and therefore reach the colon intact.30,40 Short-chain fatty acids (SCFA) are produced after the consumption of fructans and SCFA such as propionate and acetate have been shown to aid in the reduction of serum glucose and lipid concentrations.30 However, because the present review assessed changes in serum glucose concentrations, no correlations were sought between fructan consumption and serum lipid concentrations. In fact, only three RCTs in the present review involved participants who had high blood glucose levels or diabetes. Lack of evidence among those people diagnosed with diabetes indicates the need for more focused and extensive research on this population. An additional cause of lack of evidence may be due to the placebo substance used in each of the RCTs, because the placebos and fructans used may be broken down in the body by similar digestive mechanisms. If this occurred, then it may be unlikely that significant differences would be seen between those subjects consuming the fructans and those consuming the respective placebos.

The three substances used as placebos in the RCTs reviewed were maltodextrin, sucrose, and glucose. All three of these substances are rapidly digestible carbohydrates,41 although they do share slightly different chemical characteristics. Therefore, it is highly probable that different digestive mechanisms exist for the absorption of these placebos and the inulin-type fructans. The glycemic index for glucose, sucrose, and maltodextrin is 100, 58, and 163, respectively.41,42 Conversely, the glycemic index for inulin-type fructans is extremely low because, as a dietary fiber, it is not digested in the small intestine and therefore sugars are not released into the blood stream.43 The type of maltodextrin used in the previous studies is naturally linked by α-1,4-glucose linkages;44 however, resistant maltodextrin contains random 1,2-, 1,3-, and 1,4-α or β linkages that have been purposely introduced due to chemical manipulation.44 These additional linkages are not digestible by the human body and therefore this maltodextrin would act like a dietary fiber.44 However, the maltodextrin used in the studies reviewed herein was in its natural form because it was fully digestible and was not available for fermentation by the gut microflora, which are also characteristics that inulin exhibits.27

Sucrose is a disaccharide composed of glucose and fructose molecules.45 It is digested into its component sugars through a glycoside hydrolase, which catalyses the hydrolysis of sucrose. Glucose and fructose are then rapidly absorbed into the blood stream.43 The fructans used had a similar taste to that of sucrose, and physical and chemical characteristics that closely matched those of sucrose in a wide range of food applications.31 Glucose is a monosaccharide, meaning that it cannot be hydrolysed to a simpler carbohydrate45 and is thus transferred to the blood stream quickly upon consumption.43 Given that all the studies evaluated in the present systematic review were RCTs, which are the strongest type of research design,19 it is reasonable to assume that any possible differences between the placebos used and the inulin-type fructans would have been taken into account by the researchers when designing their respective studies.

Only RCTs that examined the effects of fructan consumption on fasting serum glucose values were evaluated in the present systematic review. RCTs were examined because they are considered the scientific standard in clinical research for establishing a cause-and-effect relationship.46 In each of the RCTs reviewed, the effects of inulin-type fructans on fasting serum glucose concentrations proved to be inconsistent, with no discernible pattern. There appear to be no obvious differences between study populations, in the type of fructan, or in the doses used. The wide range in the duration of individual studies, combined with the different initial characteristics of the respective study groups, such as being healthy or unhealthy, made it difficult to come to a concrete conclusion. The unhealthy population, in which decreases in glucose concentration were evident, exhibited considerable variety in terms of the symptoms of the participants. The health of the participants at the time of these studies ranged from a clinical population (diagnosis of diabetes, hypercholesterolemic, and non-alcoholic fatty liver disease) to healthy individuals.22,24,27,29

Analysis of three of the treatment groups (i.e. healthy, hypercholesterolemic or hypertriglyceridemic, and non-alcoholic fatty liver disease) did not create a clearer picture of the data. This is most likely due to the varying study lengths, age differences, and dose differences between trials. In addition, only one of the studies in the present systematic review measured both fasting and postprandial serum glucose concentrations.28 It is interesting to note that, in this study, there was a significant difference in postprandial insulin levels in the fructan-treated group. Variability between trials is not a new development in research. For example, reviews of exercise interventions have reported this variety.47 Although variability among the studies makes it difficult to arrive at a consensus, it is this same variability that indicates that the area in question is an emerging one that requires further studies to reach a consensus.

More studies concentrating on specific clinical populations, specifically a population diagnosed with diabetes, may allow for a consensus to be reached. In the three trials evaluating participants with diabetes,29–31 the age and number of participants, the dose of fructan adminstered, and the study duration varied widely. The only consensus among these three trials was the type of fructan used, which was oligofructose. Only one study29 found a significant decrease in serum glucose concentrations. In that study, participants consumed oligofructose as the fructan and consumed the lowest amount (8 g) of all the studies reviewed. In addition, the duration of that study was the shortest (14 weeks) compared with the other two studies performed on people diagnosed with diabetes. The main difference was seen in the baseline serum glucose concentrations of the study participants. In the study in which serum glucose concentrations decreased significantly,29 the participants had very high baseline serum glucose values (11 mmol/L) compared with the recommended serum glucose concentration, which is 7.00 mmol/L.2 This suggests that further research among this type of population is necessary to understand the potential long-term benefits, if any, of inulin-type fructans on lowering serum glucose concentrations.

Implications

Short-term consumption of fructans in unhealthy people diagnosed with diabetes has produced mixed results, with some studies reporting significant decreases in serum glucose concentration and other studies showing no change. However, in studies in which significant decreases were found, it is not known whether these decreases would be maintained over the longer term if fructan consumption was continued. The matter of diabetes control is an important health issue to research in Canada. The incidence of chronic diseases, such as diabetes, increases with age. Therefore, many benefits will be gained from improving the management of diabetes in the older segment of the population. Saskatchewan, in particular, has the highest proportion of older adults in Canada; hence, more research on this particular issue would greatly benefit this province. Fructans are only one of the many forms of soluble fiber and although soluble fiber can be used to control the effects of diabetes,48 the consumption of soluble fiber is only one of the many possible ways in which diabetes can be managed and and serum lipid and glucose concentrations controlled. The consumption of soluble fiber would be part of one’s overall nutritional habits. Good nutritional habits, combined with an active physical lifestyle, are the best way to manage diabetes.

Conclusions

The aim of the present systematic review was to assess the effectiveness of inulin-type fructans on fasting serum glucose concentration in humans. Based on the analysis of the studies available, it does not appear that consumption of inulin-type fructans has a significant serum glucose-lowering effect. As stated earlier, the modest number of studies available in this area could account for this finding. It has been proven through previous research that soluble fibers in general can have a serum glucose-lowering effect. This is achieved by delaying gastric emptying, which then slows the entry of glucose into the blood stream, thus decreasing the postprandial increase in serum glucose.8 This reduction in postprandial glucose is especially important for people with Type 2 diabetes. Inulin-type fructans are a type of soluble fiber and may have potential benefits in reducing postprandial glucose concentrations. More research, using an RCT design, is needed in at-risk populations, such as those with diabetes, to fully understand the effect, if any, inulin-type fructans have on serum glucose concentrations.

Disclousure

The authors declare that this article has not been, and will not be, submitted for publication in any other journal. The authors also declare that they have no conflict of interest.

Ancillary