Acute Effect of Alginate-Based Preload on Satiety Feelings, Energy Intake, and Gastric Emptying Rate in Healthy Subjects

Authors


(mmgj@life.ku.dk)

Abstract

Viscous dietary fibers such as sodium alginate extracted from brown seaweed have received much attention lately for their potential role in energy regulation through the inhibition of energy intake and increase of satiety feelings. The aim of our study was to investigate the effect on postprandial satiety feelings, energy intake, and gastric emptying rate (GER), by the paracetamol method, of two different volumes of an alginate-based preload in normal-weight subjects. In a four-way placebo-controlled, double-blind, crossover trial, 20 subjects (age: 25.9 ± 3.4 years; BMI: 23.5 ± 1.7 kg/m2) were randomly assigned to receive a 3% preload concentration of either low volume (LV; 9.9 g alginate in 330 ml) or high volume (HV; 15.0 g alginate in 500 ml) alginate-based beverage, or an iso-volume placebo beverage. The preloads were ingested 30 min before a fixed breakfast and again before an ad libitum lunch. Consumption of LV-alginate preload induced a significantly lower (8.0%) energy intake than the placebo beverage (P = 0.040) at the following lunch meal, without differences in satiety feelings or paracetamol concentrations. The HV alginate significantly increased satiety feelings (P = 0.038), reduced hunger (P = 0.042) and the feeling of prospective food consumption (P = 0.027), and reduced area under the curve (iAUC) paracetamol concentrations compared to the placebo (P = 0.05). However, only a 5.5% reduction in energy intake was observed for HV alginate (P = 0.20). Although they are somewhat contradictory, our results suggest that alginate consumption does affect satiety feelings and energy intake. However, further investigation on the volume of alginate administered is needed before inferring that this fiber has a possible role in short-term energy regulation.

Introduction

Prospective studies report that dietary fiber intake is inversely correlated with body weight gain (1,2). This is supported by human intervention studies showing weight reduction and decreased energy intake as a result of an increase in dietary fiber or fiber-supplement consumption (3,4,5). The unique properties of dietary fiber play a role in regulating energy balance as it intensifies the early satiety signals and increases or prolongs the feeling of satiety (6,7,8). The initiation of these satiety signals may be related to the ability of certain soluble fibers to delay absorption of macronutrients, thereby slowing down the postprandial release of blood glucose and insulin, and to decrease the gastric emptying rate (GER) (9).

Alginate, the major plant fiber in brown seaweed, is a glucose polymer that consists of the block-wise monomers α-l-guluronic acid (G) and β-d-mannuronic acids (M) (10). One property of alginate is its ability to gel, either in the acid environment of the stomach (pH < 3.5), or in the presence of divalent cations as an ionic gelation (11). The gelling ability, gel strength, and viscosity of alginate fibers is related to the content of G-residues (10), and may play an important role for the gastrointestinal signals that control hunger, satiety, and food intake (11,12). Currently, the literature contains only a few randomized controlled studies on the effect of alginate on appetite feelings and energy intake (11,12,13,14,15,16), and these studies show contradictory results. Two studies find no effect on satiety feelings and energy intake at all (13,14), two studies only find effects on energy intake (15,16), and two studies show a decrease in hunger feelings (11,12). Moreover, low fiber concentrations were investigated in these studies, in the range of 0.25–3.0 g administered per test day. The studies used a preload beverage as a vehicle in a small volume of 224–325 ml. In our opinion, formation of a sufficient gel mass in the stomach demands high alginate concentrations and a large amount of water for optimal hydration of the fiber. Therefore, the fiber amounts and volumes investigated in previous studies could have been too small to see a sufficient effect. Thus, we find it relevant to increase the alginate dose and volume of water in order to produce a potential basis for better enhancement of the satiety feedback mechanisms from the gastrointestinal tract.

The primary aim of the present study was therefore to test whether a higher amount of alginate in two different doses (9.9 g and 15 g), given as preloads before a meal, affects satiety feelings, GER, and energy intake in healthy subjects. A secondary aim was to assess the effect on the postprandial release of blood glucose and insulin and on blood pressure, heart rate and well-being.

Methods and Procedures

Subjects

A total of 20 healthy young Danish subjects (10 men and 10 women) were recruited through University intranet systems and advertisements on websites. Subjects were normal weight to moderately overweight (BMI 22–28 kg/m2), between 18 and 40 years of age, and free of any chronic health conditions. All subjects attended a screening visit prior to entering the study. Exclusion criteria were hemoglobin values <7.5 mmol/l, fasting blood glucose concentrations >7 mmol/l, total cholesterol concentrations >6.5 mmol/l, systolic blood pressure >140 mm Hg, and diastolic blood pressure >90 mm Hg. Subjects who smoked, were athletes (>10 h exercise/week), used regular medication (except oral contraceptives), had food allergies, or used dietary supplements were also excluded. The total number of subjects included in the trial was decided on the basis of a study of subjective appetite measurements by Flint et al. This study found that in a paired design, a minimum of 18 subjects is required in order to detect a 10% difference in mean appetite rating values with power of 80%, and with an α of 0.05 (17). Therefore, we included a total of 20 subjects, allowing for a 10% drop-out rate. Subjects meeting all the inclusion criteria were randomized and allocated to a sequence of treatment using a computer-based, randomized block design. All subjects gave their written consent after having received verbal and written information about the study. The Municipal Ethical Committee of Copenhagen and Frederiksberg approved the study (journal nr: H-C-2008-088) as being in accordance with the Helsinki II Declaration. The study was registered at http:www.clinicaltrials.gov (NCT 01101633).

Protocol

The study was designed as a randomized, double-blind, placebo-controlled, four-way crossover intervention. The preload beverage was administered at two different volumes (330 ml vs. 500 ml) and with or without alginate (3% alginate concentration equivalent to 2% total dietary fiber). The beverage was served as a preload ½ hour before a standardized breakfast and again ½ hour before the ad libitum lunch. All subjects participated in four test days at the Department of Human Nutrition (DHN) in Frederiksberg, Denmark. Each test day was separated by a >3 weeks washout period. Subjects were instructed to refrain from alcohol consumption and intense physical activity for the 24 h proceeding each test day. In the evening, before a test day the subjects consumed a standardized evening meal (prepared by kitchen staff at DHN) before 8:00 pm, after which the subjects fasted (consumption of a maximum of half liter of water was allowed).

On each test day, the subjects met at DHN in a fasting state at 8:00 am. After voiding the subjects were weighed to the nearest 0.05 kg on a decimal scale (Lindeltronic 8000, Copenhagen, Denmark), and a venflon catheter was inserted in the antecubital vein, allowing repeated blood sampling throughout the test day. Baseline measurements of blood pressure, blood samples, and visual analogue scales (VAS) scores were conducted at time point of 15 min (between 8:15 and 8:30 am). Then subjects ingested the first preload beverage at time point 0 min (with a 10 min time limit for consumption). In addition, paracetamol (1,500 mg) was taken to evaluate gastric emptying based on serum paracetamol concentrations. Immediately after consumption of the first preload blood pressure and blood samples were taken and VAS scores completed at time point 15 min (08:45 am). The standardized breakfast was served at 09:00 am (time point 30 min). After the subjects had finished breakfast, blood samples were drawn every 15 min and VAS scores, blood pressure and heart rate were measured every half hour at time points 60, 90, 120, 150, 180, and 210 min (between 09:30 am and 12:00 am). The second preload beverage was consumed at 12:00 am Immediately after consumption of the second preload, blood pressure and blood samples were taken, and VAS scores completed at time point 240 min (at 12:30 pm). The ad libitum lunch was served at 12:30 pm, and the two final blood samples and VAS score measurements were made at time points 270 min and 300 min (at 1:00 pm and 1:30 pm respectively), after the subjects had finished their lunch. Before leaving DHN at 2:00 pm, the subjects filled out a questionnaire related to gastrointestinal adverse events (AE).

Preload beverages

The sodium alginate (Algogel DPG JO; Cargill, Minneapolis, MN) was a combination of extracts from brown seaweeds Laminaria hyperborea and Lessonia Trabeculata. The preloads were given to the participants in two different volumes. The low volume (LV) was 330 ml tap water containing 9.9 g sodium alginate (= 6.3 total dietary fiber). The high volume (HV) was 500 ml tap water with 15 g sodium alginate (= 9.5 total dietary fiber), giving both volumes a 3% alginate concentration (and 2% total dietary fiber). The energy, volume, and appearance of the liquid fiber preloads were matched with the control beverage to ensure that these parameters did not influence the satiety feelings. The control and alginate beverages were artificially sweetened with aspartame (0.08 mg), colored with β-carotene (0.1% in plant oil = 0.6 g) and flavored with natural orange flavor (1.6 g) and maltodextrin. All preload beverages were prepared by mixing powder and water under stirring, and then bottled (by Adega, Denmark) before study start. All beverages were coded (A, B, C and D), labeled and stored in our metabolic kitchen at 5°C.

Viscosities [Pa·s] of the fiber preload beverages were measured using a Bohlin C-VOR rheometer (Malvern Instruments, Worcestershire, UK). The fiber preload (pH 5.50–6.00) was transferred to the rheometer cup and then left without shear for 5 min to achieve equilibrium before measurements at a constant temperature of 20°C and at 15 different shear rates ranging from 0.1 to 90 [1/s]. The measured viscosity was independent of shear rate, i.e., the liquid was Newtonian, and an average viscosity was therefore 98.4 ± 14.7 mPas for the control beverage and 480.2 ± 7.6 mPas for the alginate-based beverage.

Gelling properties of the fiber preload in acidic conditions similar to stomach pH were also tested. A fiber preload of 50 ml was stirred (600–800 rpm) while 2 ml of 0.2 mol/l hydrochloric acid was added every 10 s until pH was 1.5. The viscosity of non-Newtonian materials, like gels, is dependent on the shear rate applied during measurement. Therefore, oscillatory shear rheology was used to quantify the mechanical properties, and denoted as gel strength [Pa] of gels formed from fiber preload at pH 1.5. The oscillatory shear rheology was performed on a Bohlin C-VOR rheometer (Malvern Instruments) at a frequency of 1 Hz. During oscillations the stress amplitude was controlled and gradually increased from 1–500 Pa while the strain amplitude was recorded, and from this the material constants, elastic modulus (G'), and viscous (G”) modulus, were obtained. The measured elastic modulus was almost constant at a wide range of amplitudes, as commonly observed for many materials. An average elastic modulus [Pa] was therefore calculated and denoted as gel strength [Pa]. The gel strength for the alginate-based beverage was measured to be 1530.3 ± 16.9 [Pa]. No gel was formed from the control beverage.

Test meals

The standardized evening meal contained 4.5 MJ and consisted of pork and vegetable stew with rice, accompanied by a fruit juice (250 ml). The iso-caloric breakfast served on the test day at 09:00 am. consisted of bread, cheese, raspberry jam, yoghurt, and a glass of water (either 100 ml or 270 ml water, depending on the volume of the preload beverage, to obtain iso-volumic conditions). The iso-caloric breakfast contained 2.0 MJ in total and was to be eaten within 20 min.

The ad libitum lunch consisted of pasta with meat sauce and was served with 300 ml water. The actual amount of water consumed by the subject at the first visit (max. 300 ml) was reproduced on subsequent test days. The lunch had an energy content of 961 kJ/100 g and a macronutrient composition of 55 E% carbohydrates, 30 E% fat, and 15 E% protein. The subjects were instructed to eat until comfortably satiated and ate the meal alone, seated in special dining rooms.

Measurements of satiety feelings and palatability

VAS were used to assess subjective satiety feelings and palatability of the preloads and test meals during the test day. VAS, 100 mm in length with words anchored at each end, expressing the most positive and the most negative rating, were used to answer questions regarding satiety, hunger, fullness, prospective food consumption, thirst, well-being, and desire to eat something fatty, sweet, salty, or savory. The four standardized questions in relation to satiety feelings were: (i) How satisfied do you feel? (“completely empty” to “cannot eat another bite”); (ii) How hungry do you feel? (“not hungry at all” to “as hungry as I have ever felt”); (iii) How full do you feel? (“not full at all” to “totally full”); and (iv) How much do you think you could eat? (“not at all” to “a large amount”).VAS for palatability of the preload and test meals, had the same configuration as VAS for satiety feelings and were filled in after consumption of both preloads and meals (breakfast and lunch). The use, reproducibility and validity of the VAS have been described in Flint et al. (17).

Assessment of GER

Paracetamol was chosen as an indirect marker for GER because it is not absorbed from the stomach but from the small intestine, from where it passes into the bloodstream (18,19). The following parameters were used for identifying the rate of paracetamol absorption: peak plasma concentration (Cmax), the time to reach Cmax (tmax), and the area under the time vs. concentration curve (AUCt).

Registration of gastrointestinal AE

After consumption of the preload beverages at the end of each test day (2:00 pm), the subjects were asked to report occurrence and severity of any of the following symptoms: nausea, abdominal cramp, heartburn, feeling bloated, flatulence, and diarrhea. The scale of severity was set from 1 to 5, indicating no discomfort to very severe discomfort respectively.

Biochemical measurements

Blood samples were kept on ice, centrifuged for 10 min at 2,500g at 4°C, separated into plasma and serum, and kept at −80°C until analysis. Hemoglobin concentration was analysed using a HemoCue Hb 201+. Blood glucose concentrations in plasma were analyzed by an enzymatic colorimetric method performed in an ABX Pentra (HORIBA ABX, Montpellier, France); intra-assay coefficient of variation was 1.4%. Insulin was measured by solid-phase, 2-site chemiluminescent immunometric assay (Immulite/immuliter 1000 insulin; Diagnostic Products, Los Angeles, CA) with the use of an Immulite 1000 analyzer (Diagnostic Products). The intra-assay coefficient of variation was 2.5%. Serum paracetamol concentration was analyzed by an enzyme-specific reaction method on a Horiba ABX Pentra 400 analyzer (Cambridge Life Sciences, Cambridgeshire, UK).

Statistical methods

All descriptive data are given as means and with standard deviation (s.d.). Results are expressed as mean ± standard error of the mean (s.e.m.). The statistical analyses were performed using SAS 9.1 (SAS Institute, Cary, NC) and the level of significant difference was set at P < 0.05. Before analysis, data were tested for normality with the Shapiro—Wilk W test and homogeneity of variance, and log-transformed if necessary. Differences between treatments were tested by analysis of the mixed linear models procedure as repeated measurements with baseline level and time as covariates (RMANCOVA). Post hoc comparisons were made with Tukey—Kramer adjustment of significance levels for the pair-wise comparisons, using an unpaired t-test when the analysis indicated significant treatment effect. All VAS scores, paracetamol, glucose, and insulin responses were also calculated as the incremental area under the curve (iAUC) or over the curve (iAOC), using the trapezoidal method. Energy intake and area under the curve calculations were also tested by means of a mixed linear model procedure (ANOVA) as a 2 × 2 factorial design testing for fiber effect, volume effect and fiber × volume interaction. Relationships between VAS scores (i.e., fullness, satiety, hunger, and prospective food intake) and dose of alginate preload were tested with a Pearson correlation test. The difference between treatments in the frequencies of self-reported gastrointestinal AE was determined with the χ2 test for homogeneity.

Results

Of the 20 subjects randomized, 19 completed all four meal tests. One subject was excluded from the study due to low hemoglobin values. Subject characteristics at baseline are presented in Table 1.

Table 1.  Subject characteristics at baseline
inline image

Subjective satiety feelings and energy intake

The postprandial changes in ratings of satiety, hunger, fullness, and prospective food intake, with corresponding iAUC/iAOCs, are presented in Figure 1. We found an over effect of time (P < 0.0001) and treatment on VAS scores for satiety (P = 0.031), prospective food intake (P = 0.025), and hunger (P = 0.049). Post hoc comparisons showed the treatment effect on satiety (P = 0.038), prospective food intake (P = 0.027), hunger (P = 0.042), and fullness (P = 0.062) when treated with the HV-alginate preload compared to HV-control. This was supported by findings showing a fiber effect of decrease in hunger iAOC (−30%, P = 0.045) with a trend towards increase in fullness iAUC (+30%, P = 0.054) compared to control preloads. Interaction between fiber and volume was observed on satiety iAUC (P = 0.035) and prospective food intake iAOC (P = 0.013). Post hoc comparisons showed a greater satiety iAUC (+40%, P = 0.020) and decreased prospective food intake iAOC (−50%, P = 0.0013) for HV-alginate preload compared to HV-control. In addition we observed a possible dose-response effect as a positive correlation for increase in the feeling of satiety with administration of a higher dose of alginate fiber (R = 0.32, P = 0.05). However, no correlations with increasing fiber dose were found for hunger (R = 0.18, P = 0.27), fullness (R = 0.14, P = 0.39), and prospective intake (R = 0.21, P = 0.20). There was no significant effect of HV-alginate and LV-alginate preloads on desires to eat something sweet, salty, fatty or savory, or on well being and thirst, during the day compared to HV- and LV-control preloads respectively (P > 0.1) (data not shown).

Figure 1.

Changes in baseline-corrected VAS scores (mm). (a—d) Changes in baseline-corrected VAS satiety feelings (mm) after treatment with HV-alginate, HV-control, LV-alginate, and LV-control preloads in 19 subjects. Data are mean ± s.e.m. with corresponding iAUC/iAOC (min × mm). Data were analyzed with a mixed linear model procedure as repeated measurement with baseline level and time as covariates. Overall treatment effect on: (a) hunger (P = 0.05), (b) satiety (P = 0.03), (c) fullness (P = 0.06), and (d) prospective food intake (P = 0.03). Visual analogue scales (VAS) = 100 mm is equal to “I've never been more hungry (hunger),” “I cannot eat another bite (satiety),” “I'm totally full (fullness),” and “I really think that I can eat a large amount (prospective).” Use of a mixed linear model showed a treatment effect on hunger iAOC (*P = 0.05), satiety iAUC (*P = 0.04), fullness iAUC (P = 0.06), and prospective food intake iAOC (*P = 0.001). iAUC, area under the curve; iAOC, area over the curve; HV, high volume; LV, low volume.

We found an overall effect of fiber (P = 0.017) and volume (P = 0.06) on energy intake. Post hoc comparisons showed a significant 8% reduction in energy intake at the lunch for LV-alginate (3,237 ± 184 kJ) compared to LV-control (3,513 ± 183 kJ) preloads (P = 0.040). There was a nonsignificant numeric −5% difference on spontaneous energy intake between treatment with HV-alginate (3,141 ± 162 kJ) and HV-control (3,324 ± 162 kJ) (P = 0.20).

Palatability of the preloads and test meals

The four preloads and test meals were assessed in terms of appearance, smell, taste, aftertaste, and overall palatability (Table 2). The alginate-based preloads, independent of volume (P > 0.1), were found to be less appealing in appearance, taste, aftertaste and with lower overall palatability compared to controls when administrated before both breakfast and lunch (P < 0.001). The palatability of the standardized breakfast and ad libitum lunch was rated similarly, irrespectively of preload type (data not shown).

Table 2.  Palatability of preloads with alginate and control ingested before breakfast and lunch
inline image

GER

We found an overall time effect (P = 0.0001) but no overall treatment effect on paracetamol response (P = 0.1) irrespective of whether a high or low dose was administered. Nevertheless, we found an interaction between fiber and volume for AUCt paracetamol concentrations (P = 0.044). Post hoc comparisons showed a 15% reduction in AUCt for HV-alginate (13,874 ± 1,164) compared to HV-control preloads (16,150 ± 1,039) (P = 0.05).

No treatment effect was found on Cmax (P = 0.48) or tmax (P = 0.40). However, Cmax was numerically −14% lower for HV-alginate (101.0 ± 8.3 µmol/l) compared to HV-control (117.1 ± 6.8 µmol/l, P = 0.35). Similarly, there was a numerically longer tmax (+5.5 min) for HV-alginate (101.8 ± 14.9 min) compared to HV-control (96.3 ± 13.4 min) (P = 0.80).

Blood parameters

There was an overall effect of time (P < 0.001) and treatment on blood glucose response (P = 0.01). However, post hoc comparisons showed no difference between HV-alginate and HV-control (P = 0.10) or LV-alginate and LV-control on postprandial blood glucose levels (P = 0.94). We found an interaction between fiber and volume for iAUC glucose response (P = 0.024). Post hoc comparisons showed that the HV-alginate preload resulted in a 40% smaller iAUC glucose response (41.8 ± 10.0 mmol/l) compared to HV-control (70.3 ± 11.6 mmol/l) presented in Figure 2 (P = 0.046). No difference between LV-alginate and LV-control was observed in iAUC glucose concentration (P = 0.93).

Figure 2.

Baseline-subtracted blood glucose (mmol/l) response after ingestion of HV-alginate, HV-control, LV-alginate and LV-control preloads in 19 subjects. Data are mean ± s.e.m. with corresponding iAUC (mmol/l × mm). Model ANCOVA: effect of treatment on iAUC glucose after baseline-correction (*P = 0.01). iAUC, area under the curve; iAOC, area over the curve; HV, high volume; LV, low volume.

We found significant effect of time on insulin response (P < 0.001), but no effect of treatments (P = 0.31), regardless of volume and type of preload (data not shown).

Blood pressure and heart rate

We found a time effect on systolic and diastolic blood pressure, and heart rate (P = 0.0001). No effect of treatments was found on systolic blood pressure (P = 0.20), diastolic blood pressure (P = 0.27) or heart rate (P = 0.72), regardless of volume and type of preload (data not shown).

Self-reported AE

The LV-alginate preload tended to give a higher frequency of nausea than LV-control (P = 0.07) (Table 3). Other self-reported AE had similar frequencies after administration of all four preloads (P > 0.1).

Table 3.  Number of self-reported adverse events after ingestion of alginate or control preloads
inline image

Discussion

In this study, we tested whether different doses of alginate given as a liquid preload prior to meals could modulate satiety feelings, energy intake and gastric emptying in healthy subjects. We found that the high-dose alginate-based preload increased satiety, decreased prospective food intake and hunger, and showed a trend towards greater feeling of fullness. In addition, there was a fiber effect on decreased energy intake and a possible dose-response relationship between alginate dose and increased satiety. These results on satiety feelings are in agreement with some earlier studies, which reported a satiety effect after ingestion of alginate (11,12). However, the present findings are in contrast to results from other short-duration studies, which did not reveal any changes in satiety feelings after supplementation with alginate (13,14,16). This diverse result could be due to differences in the type of biopolymer investigated, in terms of species, M/G ratio and molecular chain lengths, which influence the viscosity of the beverage and the strength of gel mass produced in the stomach. Also differences in method of administration of the alginate, e.g., if the preload is of solid or liquid substance, could be relevant for the diverse findings in the literature. The alginate used in the present study and in the study by Hoad et al., 2004 can both be characterized as strong gelling. Generally, alginates containing a high amount of guluronate form strong gels with good heat stability, while alginates containing a high amount of mannuronate form weaker gels that are less heat stable. This can be ascribed to the selective binding of multivalent cations and the corresponding increase in gel strength increase in the order MG-blocks ≤ M-blocks < G-blocks (10). These properties might imply that guluronate, which influences gelling abilities, is important in relation to effects on satiety feelings.

An important element is whether the potential increase in satiety feelings is capable of inducing changes in energy intake. We found that the HV-alginate decreased energy intake by 5%, though this was not significant. Surprisingly, the LV-alginate significantly decreased the energy intake by 8%. A modulation of food intake after ingestion of alginate has also been found in previous studies of alginate (15,16). However, one study by Mattes et al., (13) did not reveal any effect on energy intake after ingestion of an alginate-based breakfast bar. Nevertheless, the method used was diet recall, which is less precise than determining changes in the amount of food consumption of an ad libitum test meal. A possible explanation for the lack of significant effect from the HV-alginate preload on energy intake could be that our preload is solely dependent on gastric acid for gel formation. The greater volume of water of the HV-alginate preload, compared to LV-alginate preload, may have diluted the gastric acid content in the stomach and thereby inhibited complete gel mass formation. In addition, the relatively large volume of water in the HV-control preload beverage may have acted as a positive control. Dennis et al., found that premeal water consumption (500 ml) facilitated an acute reduction in test meal energy intake in overweight subjects compared to no water ingestion (20).

The decrease in energy intake of our alginate preloads vs. placebo of ∼183 kJ is modest, though perhaps on a sustained basis this energy gap might have some relevance for regulating the energy balance. On the basis of data from the US National Health and Nutrition Examination Survey, Hill et al., estimated that a reduction in energy intake by ∼400 kJ/day would offset weight gain in 90% of the population (21). The observed reduction in energy intake in our study makes a contribution similar to these findings, which could be important at overall population level, though much more so if the effect could be extrapolated to 270 kJ/day from administrating preloads before three main meals. Still, future studies will need to address this more directly.

The effects on satiety and food intake observed in this study may be explained by multiple mechanisms, with gastrointestinal signals being crucial factors. An indicator for the relevance of these signals could be linked to the observed difference in the response in GER and glycemic control between alginate and control preloads. However, our findings were contradictory in the sense that we observed indications for decreased GER for HV-alginate, but not for LV-alginate preload. The contradictory GER results could be due to the paracetamol method itself. There are a number of limitations to the paracetamol method, as it only provides an indirect estimation and therefore does not represent an optimal measure of gastric emptying, as e.g., scintigraphy (18,22). Moreover, we do not know how paracetamol itself affects gastric response. It could also be that the drug is trapped in the gel mass and thereby interferes with the absorption rate as we did not observe any difference in tmax.

One previous study investigating the effect on GER, found delayed gastric emptying, measured by aspirated radioactive gastric content, after alginate administration (23). However, two interventions using the well-validated scintigraphy method found that gastric emptying did not change after ingestion of alginate fiber (11,14), indicating that the effect of alginate fiber consumption on subsequent gastric emptying it is still unclear.

One mechanism on the postprandial satiety feelings could be an increase in gastric distension and activation of the satiety signals, through the vagal and splanchnic nerve and from activation of the chemo- and mechanoreceptors in the stomach wall (24,25) governed by the gastric gel formation of alginate.

Other possible mechanisms related to the gastrointestinal tract upon alginate consumption could be a potential increase in the viscosity of the intestinal contents when the alginate gel lumps leave the stomach. In relation to this, soluble fibers may slow the rate of nutrient absorption, as the unstirred water layer at the luminal surface is thickened. This effect could delay the removal of nutrients from the intestinal lumen, resulting in exposure of nutrients further down the small intestine (28) and leading to pronounced inhibitory feedback on appetite from the release of gut peptides such as glucagon-like peptide 1 and peptide YY (26,27). The latter exposure of intestinal mucosa in the ileum to food constituents influences the activation of the ileal brake. The ileal brake is a negative feedback mechanism that inhibits gastric emptying and small intestinal transit, which could have potent anorexic effects in humans (26,27).

It has also been reported that some viscous fibers may reduce the activity of certain digestive enzymes (e.g., pepsin, trypsin) (27) which could have beneficial effects on glycemic control at a subsequent meal (29).

We found that the iAUC for glucose was reduced by 40% after HV-alginate preload ingestion. Similar results have been seen in earlier meal-test studies where the postprandial incremental blood glucose response of healthy subjects was improved after supplementation with alginate or guar gum fiber (30,31). Furthermore, Torsdottir et al., (23) found that alginate supplementation in humans with diabetes improved postprandial glycemia and reduced the rise in insulin response. However, the present data do not support this finding, in that we found HV-alginate preload improved the glycemic response without an effect on the insulin response.

Postprandial blood glucose concentration is determined by rates of glucose appearance and clearance. The possible mechanism behind lower blood glucose levels is similar to that behind the enhancing effect on satiety. Soluble fibers can slow gastric emptying by forming a gel matrix, increasing viscosity in the small intestine, and delaying the digestion of food due to a decreased diffusion of nutrients for absorption (thickened unstirred water layer) and contact between food and digestive enzymes (32,33).

Highly viscous beverages and the addition of soluble fibers to diets are sometimes related to a reduction in palatability, which could be argued to have a confounding effect on subsequent food intake (9,34,35). Studies measuring palatability of alginate supplementation often find a decreased palatability of the alginate product (11,13,30). When we developed the preloads it was our aim to match the sensory properties. Unfortunately, we were not successful in doing this, as there was found a significant difference between the sensory properties of the alginate-based preload and the control. In addition to measuring palatability, this could also be seen as a limitation of our study design as we did not examine opinions of mouth feel, preload texture or whether there were any difficulties with ingesting the preload. As the alginate preload was more viscous than the control, one could also expect a difference in ratings of texture or mouth feel. However, there was no significant difference in the taste evaluation of the subsequent lunch, consumed 30 min after preload, for any of the four preload types. Therefore it could be assumed that differences in energy intake between treatments were not only attributable to differences in the palatability of the preloads. Zijlstra and colleagues did not see this after two preloads which differed in viscosity but had identical nutrient composition (36), thus we believe that the alginate and not viscosity per se would be the main reason for the observed differences in energy intake observed in the current study.

High fiber concentration supplements have been associated with possible gastrointestinal AE (34,37). However, no differences in subjective gastrointestinal tolerance symptoms were observed in our meal test. This is in line with earlier studies (13,16,30,31) on alginate, which also found rating of symptoms of gastrointestinal AE to be similar between alginate fiber and control products. It has previously been suggested that the alginate fiber is more slowly fermented than other viscous fibers (29,31), which might make alginate a more tolerated fiber source than rapidly fermented fibers which could be associated with greater discomfort in the gastrointestinal region.

There are some limitations to the present study that warrant further investigation into the effect of alginate on satiety feelings and energy intake. There is evidence that the menstrual cycle has confounding effects on gastrointestinal function, appetite feelings and energy intake (38), and should be controlled for when evaluating a preload paradigm. We did not completely track which phases of the menstrual cycle the female subjects were in. However, all female subjects used oral contraceptives, and visits were not scheduled in self-reported menstruating weeks. In addition, this study assessed only effects in the short-term, and any effects observed in such acute studies may not be indicative of energy balance in the long term. Finally, future studies investigating GER will need to be more accurate when applying the paracetamol method. Possibly, by including correction factors (39).

In conclusion, our results showed that the high-dose alginate-based preload increased satiety, and decreased feelings of prospective food intake and hunger, indicating enhanced perceived short-term satiety feelings. Besides this, the high-dose alginate seemed to blunt AUCt paracetamol and iAUC glycemic response, suggesting some alteration in the absorption rate. The dose inconsistency of the effect from alginate preloads on energy intake compared to controls, however, warrants further studies. These should investigate alginate dose and volume of vehicle for optimal appetite and intake control, and which of these parameters is most important.

Acknowledgment

M.G.J., M.K., A.B., and A.A. designed the study. M.G.J. wrote the protocol and collected the data. M.G.J., M.K., A.A., J.C.K. analyzed the data, and all authors participated in the discussion of the results and commented on the manuscript. The study was supported by grants from S-Biotek Holding ApS, Denmark, and FOOD Research School/LIFE, University of Copenhagen, Denmark. The alginate used in the study was provided by Cargill A/S, Denmark. The authors are grateful to the subjects participating in the study, to the laboratory technician J. Jørgensen for her technical assistance, and to the staff at the metabolic kitchen at the Department, C. Kostecki and Y. Fatum, for assistance with test meals and preloads. ClinicalTrials.gov (NCT 01101633).

Disclosure

The authors declared no conflict of interest.

Ancillary