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

Disclosures

Nutr Metab. 2015;12(36) 

In This Article

Background

Lifestyle interventions can significantly reduce the risk of developing type 2 diabetes (T2DM)[1] with weight loss being the primary mediator of the reduction in risk.[2] However, outside of a labour-intensive clinical trial setting, long-term weight loss is notoriously difficult to achieve.[3,4]

One explanation for the rarity of successful weight loss maintenance is that an energy deficit and loss of body fat are both associated with increases in appetite and food intake.[5,6] Therefore, interventions such as glucagon-like peptide-1 (GLP-1) agonists aimed at decreasing or ameliorating such changes in appetite help promote long-term weight loss maintenance.[7] Non-digestible carbohydrate (dietary fibre) has long been linked to reduced food intake,[8] and fermentable carbohydrates (FCHO) may be particularly effective.[8] This class of carbohydrate passes undigested and unabsorbed from the upper gastrointestinal tract to the colon where bacterial fermentation produces short chain fatty acids (SCFA) that can stimulate GLP-1, and regulate other appetite hormones.[9] Furthermore, as a natural dietary compound, FCHOs could represent a widely applicable public health intervention. By promoting a natural reduction in appetite, less frequent clinical support may needed. In a year-long study in overweight adolescents, there was a significantly lower rise in body mass index (BMI) in the inulin group, without receiving input from health professionals.[10]

A key mediator of the beneficial effect of weight loss on insulin sensitivity is the loss of ectopic fat,[11] which is highly correlated to insulin resistance and T2DM.[11–13] Intriguingly, FCHO has been shown to reduce ectopic fat in animal studies independent of weight loss.[14] This raises the possibility that this natural dietary carbohydrate may not only enhance weight loss efforts, but may also promote loss of this metabolically-deleterious fat in humans, even after accounting for weight loss. In healthy and insulin-resistant subjects, FCHO has been shown to improve insulin sensitivity,[15,16] but it is not clear whether a reduction in ectopic fat plays a role. No previous study has examined the effect of a dietary intervention on ectopic fat in subjects with prediabetes.

Here we report the effect of 30 g/day (following a 4-week dose escalation period) inulin supplementation taken alongside a 9-week weight loss program and a 9-week weight maintenance period. We hypothesise the inulin will result in greater weight loss maintenance at the 18-week follow-up in subjects with prediabetes, and will promote a reduction in ectopic fat and insulin resistance independent of weight loss assessed by using the gold-standard magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) techniques.

Volunteers

Subjects with known prediabetes or high risk factors for prediabetes were identified from local GPs registers. Informed consent was obtained prior to the study. The protocols were approved by the North West 1 Research Ethics Committee (registration number: 10/H0717/32) (Clinical trial number: NCT01841073) and conformed to the Declaration of Helsinki. An oral glucose tolerance test (OGTT) was performed to clarify glycaemic status.[17] Inclusion criteria were age ≥ 18 years, BMI of 25–35 kg/m2, prediabetes (impaired fasting glucose (IFG) and/or impaired glucose tolerance (IGT)) using the American Diabetes Association criteria[17] and a stable body weight for 3 months prior to the study. Exclusion criteria were normal glucose tolerance, T2DM, gastrointestinal disorders, pregnancy or breastfeeding, prescribed medication that affects appetite or glucose homeostasis and consumption of prebiotic products or antibiotic use within 3 months of the study start date. Volunteers were randomised by BMI and gender using a random-number table, with an allocation ratio of 1:1.

Experimental Design

The study was a double-blinded randomized parallel control trial examining the effect of 30 g/day inulin (Synergy1, dOrafti, Tienen, Belgium) versus the control cellulose (Vitacel® Powdered Cellulose L 600–20, J. Rettenmaier & Söhne GmbH + Co.) alongside a 9-week weight loss and 9-week weight maintenance program (Fig. 1). The fibre cellulose was chosen as a comparator as this unbranched polymer of glucose molecules linked by (β4-1) bonds undergoes minimal colonic fermentation. The 30 g dose was chosen based on previous studies by our group.[18,19] Both supplements were given as 10 g sachets taken three times a day with food or drink. To reduce potential gastrointestinal side-effects supplements were increased by 10 g/day every 2 weeks to reach the 30 g/day dose. Therefore, by week five of the nine-week weight loss phase, all subjects were on the maximum 30 g/day dose. The inulin and cellulose sachets were assigned letter A or B and were otherwise identical.

Figure 1.

Schematic showing study outline, including the timings of blood samples, VAS and breath hydrogen measure during the MTT. H2: breath hydrogen measure; MTT: meal tolerance test; MRS: magnetic resonance imaging; MRS: magnetic resonance spectroscopy; VAS: visual analogue scales

During the 9-week weight loss program each subject underwent a standardized dietary intervention comprising four dietary sessions 2–3 weeks apart with a registered dietitian (Fig. 1). The dietitian was blinded to the supplement allocation. Energy requirements and assigned portion sizes for starches, protein, milk and dairy, fats and sugars and fruits and vegetables were determined using a ready-reckoner developed at Hammersmith Hospital. Following the 9-week visit, subjects were given no further input or support from the study team. They were merely asked to try and maintain the weight that they had lost.

All subjects attended a study day at baseline, 9 weeks (representing the end of the weight loss phase) and 18 weeks (end of weight maintenance phase) for a meal tolerance test (MTT) (Ensure Plus™ (220 ml), Total energy: 1380 kJ, 44.4 g carbohydrate, 10.8 g fat, 13.8 g protein). Blood samples were taken at −15, 0, 5, 10, 15, 20, 30, 45, 60, 90, 120, 150, 180 min for measurement of glucose, insulin and GLP-1 (Fig. 1).

Appetite was assessed during the MTT using validated[20] visual analogue scales (VAS) at frequent intervals, followed by an ad libitum meal test.[20] The appetite questions used were "How hungry are you right now?" (general hunger), "How much do you think you could eat right now?" (prospective food consumption), "How pleasant would it be to eat right now?" (desire to eat) and "How full do you feel right now?" (feeling of fullness in the stomach) (Additional file 1 http://static-content.springer.com/esm/art%3A10.1186%2Fs12986-015-0033-2/MediaObjects/12986_2015_33_MOESM1_ESM.docx).

Body composition was assessed by a 9-electrode bioelectrical impedance (BIA) device (Tanita BC-418 MA Segmental Body Composition Analyzer) (Tanita Corporation, Tokyo, Japan). The BIA device provides estimates of % body fat (±0.1 %), fat mass (±0.1 kg), and free fat mass (±0.1 kg). In addition, a subset of subjects who were eligible (no metal in situ or reported claustrophobia) (n = 20) underwent MRI and MRS to assess total and regional fat volumes at baseline, week nine and week 18. Rapid T1-weighted magnetic resonance images were obtained using a 1.5 T Phillips Achiva scanner (Phillips, Best, the Netherlands)[21] with intrahepatocellular lipid (IHCL), and intramyocellular lipid in the soleus and tibilalis muscles (IMCL-S and IMCL-T, respectively) assessed by MRS as previously described.[22]

Compliance was assessed by asking subjects to return unused sachets and breath hydrogen was measured as a proxy of colonic fermentation using a breath hydrogen monitor (Gastrolyzer, Bedfont Scientific Ltd. Kent, UK).

Laboratory Analysis

Glucose was collected into fluoride oxalate tubes and measured using an Architect ci8200 analyzer (Abbott Diagnostics, Maidenhead, UK), with an assay detection limit of 0.3 mmol/L and an intra-assay coefficient of variation (CV) of 1 %. Plasma insulin samples were collected into serum gel separator tubes containing gel clotting activator Vacutte® and measured with a commercial radioimmunoassay kit (Millipore; Watford, UK). The sensitivity and intra-assay CV for insulin were 7.1 pmol, and 3.0 % respectively. GLP-1 samples were collected into lithium heparin tubes, with aprotinin (Trasylol, Bayer, Newbury, UK) (200 μL/7.5 ml blood) added. Blood samples were spun at 4000 g at 4 °C for 10 mins, separated into plasma and stored at −20 °C until analysed using an in-house radioimmunoassay.[23] The sensitivity and intra-assay CV of the GLP-1 assay were 7.5 pmol/l and 3.3 % respectively.

Calculations and Statistical Analysis

Based on a study by Parnell et al.[24] with an expected weight loss of 2.0 kg and a standard deviation of 2.0 kg based on 0.8 power to detect a significant difference (P < 0.05, 2-sided) we estimated a minimum of 32 subjects were needed; 16 additional subjects were added to account for dropouts. Data are presented as mean ± standard error of the mean (SEM) for normally distributed data and median and interquartile range for non-normally distributed variables. The postprandial response curves for GLP-1, glucose and insulin were calculated as total area under the curve (tAUC) using the trapezoid rule. The primary outcome of weight change between the inulin and cellulose groups was calculated using an ANCOVA with baseline weight as a covariate. The delta change between the insulin and cellulose groups (between group difference) for glucose, inulin, GLP-1, and measures of adiposity and ectopic fat were compared using ANCOVAs, with change in weight as a covariate. Non-parametric tests were used for variables not normally distributed. A p value of less than 0.05 was considered significant. Analyses were performed using GraphPad Prism Version 5.0 (GraphPad Software, San Diego, CA) and ANCOVA was performed using SPSS 20.0 (SPSS Inc. Chicago, IL USA). The homeostatic model assessment of insulin resistance (HOMA-IR) and Matsuda index were used to measure fasting and postprandial insulin sensitivity.[25,26]

processing....