A Randomized Controlled Trial

The Effect of Inulin on Weight Management and Ectopic Fat in Subjects With Prediabetes

Nicola D. Guess; Anne Dornhorst; Nick Oliver; Jimmy D. Bell; E. Louise Thomas; Gary S. Frost


Nutr Metab. 2015;12(36) 

In This Article


Weight loss reduces the risk of diabetes, and medications which reduce appetite can promote long-term weight loss.[7] Previous studies have shown that FCHO reduces appetite, food intake and body weight, independent of lifestyle change.[10,18,24] We extend these findings to show that the consumption of inulin enhances a traditional calorie-restricted lifestyle program. An added benefit of the inulin supplement was a greater reduction in intrahepatocellular and intramyocellular lipid in the soleus muscle even after accounting for weight lost.

Expert bodies recognize the importance of frequent visits in primary care (more than once a month, face-to-face contact during the first three months) in achieving long-term weight loss maintenance.[27] However, the labour-intensive nature of this approach reduces its widespread applicability. Therefore an important component of our study design was to ensure that the subjects would not receive dietary advice or support during the final 2 months of the study. Our finding that the inulin group lost more weight between weeks 9–18 compared to cellulose therefore suggests that inulin may be a useful adjunct to traditional lifestyle approaches to diabetes prevention. Furthermore, despite the 5 % weight loss at week nine, subjects taking inulin ate significantly less (~270 kcal less, p = 0.027) at the follow-up ad libitum meal, with no consequent rebound in food intake at the 18 week visit despite a total 7 % weight loss in the inulin group, suggesting that inulin's effect on weight management is mediated via appetite modulation.

Unexpectedly, the cellulose group also continued to lose weight during the weight maintenance period, albeit a modest amount. Firstly, the length of the study was likely not long-enough to measure true long-term weight loss maintenance, and both groups may have simply continued with the dietary changes they had made during the 9-week weight loss period. Weight loss tends to plateau at 6 months, after which weight exhibits a gradual but continuous rise.[28] We had selected the 9-week weight loss and 9-week weight maintenance periods as a balance between degree of participant burden and feasibility.[29,30] Our findings at this time therefore support a role for inulin in weight loss, as opposed to weight-loss maintenance. We suggest that a longer follow-up period should be incorporated in future study designs.

Secondly, we had selected cellulose as the control as it is known to be the least fermentable of the fibres.[31] However, adaptations of the gut microbiota over time have also been reported following cellulose ingestion[32] and cereal fibre (24 g per day) has been shown to increase GLP-1 secretion taken over a year.[33] We suggest that more work should be carried out to understand the physiochemical properties of dietary fibres and their relationship to weight management. Nevertheless, since dietary fibre is inversely associated with BMI[8,34] the use of cellulose as a control group would suggest our data underestimate the likely magnitude of effect of inulin on weight loss.

Inulin consumption not only promoted weight loss but was also associated with greater loss of triglyceride in the liver (IHCL) and soleus muscle (IMCL-S), even after controlling for differences in weight lost. We were able to collect MRI and MRS data on 19 subjects, and similar studies examining, IHCL, IMCL and adiposity have used comparable sample sizes.[35,36] We therefore extend findings of a beneficial effect of FCHO on ectopic fat from animal studies,[14] and from human studies in non-alcoholic fatty liver disease[37] to suggest that FCHO is also able to reduce IHCL and IMCL-S content in subjects at risk of diabetes independent of weight loss. A previous study has shown that 12 weeks' supplementation with the FCHO oligofructose led to greater weight loss as fat as measured by x-ray absorptiometry.[24] However, our data on the proportion of weight lost as fat are inconclusive, given the lack of significance in the MRI data and the caution required in interpreting bioimpedance measurements.

The reduction in IHCL following inulin supplementation is of particular importance given the potent and consistent relationship between triglyceride content of the liver and metabolic disease, particularly diabetes.[11,13] Furthermore, previous studies have found an increase in lipid content of the soleus muscle but not the tibialis in the off-spring of subjects with type II diabetes[12] and Indian males with diabetes have significantly higher IMCL-content in the soleus compared to healthy controls.[38] The greater oxidative muscle fibre content of the soleus muscle may be related to insulin resistance due to altered insulin-dependent glucose transport.[39]

The mechanism by which FCHO reduces fat deposition appears to be related to increased fat oxidation. Rats fed a diet high in viscous (fermentable) fibres have increased expression of carnitine palmitoyltransferase 1B (CPT-1B) – the rate-limiting enzyme in the soleus muscle, alongside other key fat oxidation genes: peroxisome proliferator-activated receptors 1α and δ, uncoupling protein 3, and citrate synthase.[40] In humans, fat oxidation is increased 3 h after consumption of a low glycaemic index breakfast, indicating that fermentation of unavailable carbohydrates may influence oxidation in vivo.[41] It has also recently been shown that inulin down-regulates hepatic lipid production alongside increases in portal SCFA concentrations in mice fed a high-fat diet.[42]

Despite the effect of inulin on liver fat we did not observe any added effect on insulin sensitivity after accounting for weight loss. Dietary studies on the effects of FCHO on insulin sensitivity have shown improvements in peripheral and/or hepatic insulin sensitivity.[15,16,43] We did not find any changes in fasting (HOMA-IR) or postprandial insulin sensitivity (Matsuda Index) in this study. However, the distinct pathophysiology which underpins the different prediabetic states may have been a confounding factor. Subjects with IFG are known to have hepatic insulin resistance, but normal peripheral insulin sensitivity;[44] while subjects with IGT are known to have normal or mild hepatic, but marked peripheral insulin resistance.[44] It is not known whether inulin affects hepatic or peripheral insulin sensitivity, and this study was not powered to examine differences between the prediabetic states. This should be considered in future studies. Nevertheless, there was a significant reduction in FPG after controlling for weight loss in the inulin group at week nine. It is unclear why this was no longer different at week 18, but insoluble fibres have shown repeated inverse associations with diabetes risk in cohort studies,[45] with the mechanism currently unknown. Potentially, the fermentation of cellulose and the rise in GLP-1 may have played a role. Furthermore, other changes made in the diet in both the inulin and cellulose groups may also have contributed, as all subjects were encouraged to eat five or more servings of fruit and vegetables daily and increase their consumption of high-fibre foods. Total dietary fibre is known to be inversely related to a number of cardiovascular risk factors;[46] however our data point to specific metabolic effects of fermentable fibres which may be of particular benefit to individuals at risk of diabetes. Use of stool analysis in future studies would be very useful to help delineate some of these effects.

We acknowledge some limitations in the study. While compliance in dietary studies is always challenging, the breath hydrogen significantly increased when the supplement was being taken, providing a degree of dietary compliance. The high return rate of the used sachets also suggests that the effects seen in this study were due to chronic consumption of the supplementation. We also did not measure physical activity which is known to affect insulin sensitivity and loss of ectopic fat. We therefore cannot rule out that exercise may have influenced the results seen here, and activity meters should be used in future studies. Finally, there are known gender differences in body composition.[47] While the study did not aim to determine gender differences in the study outcomes, this is certainly something that should be examined in a study powered to do so.