Previous studies have demonstrated the satiating properties of soups compared with solids; however, the mechanisms controlling soup-induced satiety are unknown. This study aimed to understand the physiological mechanisms causing soup to be more satiating.
A total of 12 volunteers were tested on three occasions after a solid meal, chunky soup or smooth soup test meal for gastric emptying (GE) using the sodium [1-13C] acetate breath test, satiety using visual analog scales (VAS) and glycaemic response (GR) using finger prick blood samples.
There was a significant difference in GE half-time (P=0.022) and GE ascension time (P=0.018), with the longest GE times for the smooth soup and the shortest for the solid meal. The GR area under the curve was significantly different between meals (P=0.040). The smooth soup had the greatest GR (87.0±49.5 mmol/l/min), followed by the chunky soup (65.4±48.0 mmol/l/min), with the solid meal having the lowest GR (61.6±36.8 mmol/l/min). Volunteers were fuller after the smooth soup compared with solid meal (P=0.034).
The smooth soup induced greater fullness compared with the solid meal because of a combination of delayed GE leading to feelings of gastric distension and rapid accessibility of nutrients causing a greater glycaemic response.
Obesity and its related complications have become an overwhelming problem in our society. In 2009, obesity was estimated at ∼23% among British adults, and with rates of obesity consistently increasing, ways in which obesity can be prevented are paramount in helping to thwart the increase in obesity rates into the future.1 Reducing energy intake through increasing satiety may help to prevent obesity. One factor that has been identified as a major influence in the control of satiety is the physical state of a food.2
In general, solids are considered to be more satiating than liquids,3, 4, 5, 6 perhaps because of their lack of orosensory stimulation;7 however, this is still a subject of debate.8 However, soups are a different phenomenon in that they can cause stronger satiety sensations than solid foods. Explanations have included the fact that soups have a low energy density and high water content, and hence the greater the volume ingested, the fuller the stomach becomes, signalling satiety via the vagus nerve.9 This may have been the case in some older studies; however, recent research has controlled for water volume,10 yet soup is still more satiating than a solid meal. Previous research on the satiating properties of soups has not examined the mechanisms controlling soup-induced satiety. It is believed that the physical state of food and particle size affect the gastric emptying (GE) rate and glycaemic responses (GR). The rate of GE will affect both the gastric distension signalling fullness via the vagus nerve9 and the delivery of nutrients into the duodenum initiating satiety via release of satiety hormones.11 GE and GR are hence both strongly related with a slower GE leading to a reduced GR. The lower the GR to a food the more satiating the food is considered to be via the glucostatic theory.12, 13 The purpose of the study is to understand how the physiological mechanisms of GE and GR cause soup to be more satiating. This will be achieved through the measurement of GE, GR and satiety following the consumption of three meals—a solid meal, a chunky soup and a smooth soup.
Materials and methods
A total of 12 healthy participants (6 male and 6 female; 28.7±5.9 years; 1.73±0.09 m; 68.7±14.0 kg; body mass index male: 23.5±2.9 female: 22.1±2.8 kg/m2) were recruited for the study. Before inclusion in the study, potential participants were briefed regarding all aspects of the experiment and were given the opportunity to ask questions. This was followed by a health assessment, which included anthropometric measurements and a health questionnaire (giving details of food allergies/intolerances, metabolic diseases, special dietary needs and smoking habits and time of the menstrual cycle). Those who fulfilled all the acceptable criteria (body mass index: 18.5–24.99 kg/m2; blood pressure: 110–120/75–85 mm Hg; age: 18–35 years; fasting blood glucose: 4–6 mmol/l; not on prescription medication; nonsmoking; no genetic or metabolic diseases) were included in the study. Physical activity was quantified using Baecke’s questionnaire14 and only those not partaking in competitive sports and endurance events were included. Subjects were requested to come to the laboratory between 0700 and 0830 h after a 12-h overnight fast.
Potential participants were informed about all aspects of the experiment and gave written informed consent. The study was conducted according to the guidelines laid down in the Declaration of Helsinki, and was approved by the Oxford Brookes University research ethics committee.
A randomized, crossover, within-participant, repeated-measures, nonblind design was adopted. The experiment consisted of three different treatments and participants came into the laboratory on 3 different days, separated by at least 1 day. Here they ate one of the three different test meals and had their GR, GE and satiety measured. On the day before testing, participants were asked to not consume alcohol and caffeine-containing drinks and to restrict their physical activity levels.
Study day protocol
On arrival in the laboratory, volunteers had baseline breath samples for GE, finger prick blood samples for glucose and satiety measurements taken. They were then presented with one of the three test meals and 15 min in which to consume it. For the next 3 h, volunteers had their GE, GR and satiety measured.
The protocol used to measure the GR was adopted from that described by Brouns et al.15 and is in line with procedures recommended by the FAO/WHO.16 Blood was obtained by finger prick using the Unistik 3 single-use lancing device (Owen Mumford, Woodstock, UK). Before a finger prick, subjects were encouraged to warm their hand to increase blood flow. In order to minimize plasma dilution, fingertips were not squeezed to extract blood but were instead gently massaged starting from the base of the hand moving toward the tips. The first two drops of expressed blood were discarded and the next drop was used for testing. Blood glucose was measured using the HemoCue 201+ Glucose analyzer (HemoCue Ltd, Dronfield, UK). The HemoCue is a reliable method of blood glucose analysis.17 To ensure accuracy of data, the HemoCue instruments used in the study were compared against a blood glucose analyzer (YSI 2300 stat, YSI Inc., Yellow Springs, OH, USA) to ensure that samples of measurements were within 5% of each other. Fasting blood samples were taken at −5 and 0 min, and the test food was consumed immediately afterwards. The participants consumed the test food and the water at a comfortable pace, within 15 min. Further blood samples were then taken at 15, 30, 45, 60, 90 and 120 min.
The GR data were converted to ‘the change in GR’ values. The ‘change in GR’ was calculated by computing the difference between the blood glucose concentration at a time point and mean baseline blood glucose concentration (based on two baseline values taken 5 min apart (−5 and 0 min). As it represented the relative increment in the GR at any time point compared with the baseline value, it was this ‘change in GR’ that was used for all further analyses, including incremental area under the curve (IAUC; calculated using the trapezoidal rule),15, 18 blood glucose response curve construction and statistics.
Sodium salt of 1-[13C] acetate was used to measure GE as acetate is hydrophilic, poorly absorbed in the stomach and rapidly metabolized after absorption. Sodium [1-13C] acetate is considered a reliable and valid method for identifying changes in GE of semisolids.19
Breath samples were collected by blowing gently into a 10 ml Exetainer (Labco, Buckinghamshire, UK) with a drinking straw and replacing the cap just before the end of exhalation. Breath samples were collected at baseline and every 15 min postprandially until 4 h. Breath samples were analyzed using isotope ratio mass spectrometry (DeltaV, ThermoFisher Scientific, Hemel Hempstead, UK) and results were expressed relative to V-PDB, an international standard for known 13C composition. The 13CO2 values were expressed as the excess amount in the breath above baseline and converted into moles. Data are then displayed as percentage of 13CO2 dose recovered per hour and cumulative percentage 13CO2 recovered over time. CO2 production was assumed to be 300 mmol/m2 body surface area/h. Body surface area was calculated using a validated weight/height formula.20 This was then fitted to a GE model developed by Ghoos et al.21 For all the data, r2 coefficient between the modeled and raw data was calculated and r2>0.95. From this model several parameters were measured. Lag phase and half-time were calculated using the formulae derived by Ghoos et al.21 The lag phase is the time taken to maximal rate of 13CO2 excretion22 and is equivalent to the time of the inflection point.23, 24 The half-time is the time it takes 50% of the 13C dose to be excreted.22 The latency phase23 is the point of intersection of the tangent at the inflection point of the 13CO2 excretion curve representing an initial delay in the excretion curve. The ascension time23 is the time course between the latency phase and half-time, representing a period of high 13CO2 excretion rates.
One hundred millimeter continuous line visual analog scales (VAS) were utilized to measure subjective feelings of hunger, fullness, desire to eat and prospective food consumption, at baseline (0 min) and then every 15 min for the first hour and every 30 min for the following 3 h after the commencement of eating the test food. The VAS ratings were quantified by measuring in millimeters, the distance between the left end of the scale and the point marked by the participant. The ‘change in the subjective feeling’ was calculated by computing the difference between the response at a time point and the baseline value (at 0 min). Using the ‘change in subjective feeling’ data, temporal curves were constructed for each of the four VAS questions for the testing time. The IAUC (using the trapezoidal rule) was then calculated for each of these curves.
The test meals were based on the meal used by Rolls et al.25 Each test meal had the same constituent ingredients: 51.8 g rice, 69.4 g chicken, 40.8 g carrot, 44.3 g peas, 19.0 g onion, 26.7 g mushrooms and 22.1 g celery. However, each of the three meals was presented in a different form. The first test meal (solid meal) was presented as a solid test meal consisting of rice, vegetables and chicken served alongside 400 ml of water. The second test meal (chunky soup) had the vegetables, chicken and 250 ml of water blended together to form a liquid, and the rice was then added to the liquid to make a chunky soup. This was served with 150 ml of water, to make up 400 ml in total. The final test (smooth soup) meal had all of the ingredients blended together with 350 ml of water to form a smooth soup. This was served with 50 ml of water to make up the 400 ml.
The rice was cooked separately in water containing 100 mg of sodium [1-13C] acetate for measurement of GE. The volume of water used was calculated so that there was no water remaining after cooking. The rice utilized was a pure white basmati of known varietal purity (Oryza sativa L.; strain HBC-19) and was from a single cultivated batch. Each meal contained 50 g of available carbohydrate.
Studies on assessment of GR have been based on 10 volunteers.26 A sample size of 12 was utilized to account for any dropouts.
Statistical analysis was conducted using SPSS version 19 (SPSS Inc., Chicago, IL, USA). A Kolmogorov–Smirnov test indicated that the data were normally distributed. Data are presented as mean±s.d. Differences in the GR and satiety IAUCs as well as the GE parameters were compared using one-way repeated-measures analysis of variance (P0.05). Comparisons between meals were done by examining contrasts within the analysis of variance with Bonferroni correction. Comparisons in satiety and GR across time were compared using two-way repeated-measures analysis of variance with two within-subject factors—time and meal.
There was a significant difference in GR across time for the three test meals (P<0.001). There were significant differences in GR between meals (P=0.028) and there was a significant meal–time interaction (P<0.001; Figure 1). The major difference in GR between the meals was between the smooth soup and the solid meal (P=0.038) and approaching significance between the smooth soup and the chunky soup (P=0.053).
There was a significant difference in GR IAUC between the three meals, with the smooth soup having the highest GR IAUC and the solid meal the lowest (P=0.040; smooth soup 87.0±49.5; chunky soup 65.4±48.0; solid meal 61.6±36.8 mmol/l/min).
There was a significant difference in GE half-time (P=0.022) and ascension time (P=0.018) that primarily existed between the solid meal and the smooth soup (P=0.022 for both). There were no differences in other GE parameters. The longest GE times were for the smooth soup and the shortest for the solid meal (Table 1).
There was a significant difference across time in all of the satiety parameters (P<0.001). There was a significant time–meal interaction for all satiety parameters (P<0.05). There was a difference between the solid meal and smooth soup for fullness (P=0.034; Figure 2), with the smooth soup causing greater feelings of fullness. There was also a difference in fullness IAUC between the solid meal and smooth soup (P=0.033; Table 2), with the smooth soup inducing the greatest fullness. There were no differences in the other satiety IAUC parameters.
Many studies have examined the effect of soups on satiety;2, 10, 25, 27 however, research investigating why soups may be more satiating is relatively unexplored. The current study showed that after consumption of soup, fullness is increased, GE is delayed and GR increased compared with an isoenergetic and isovolumetric solid meal.
The present study is in agreement with many studies that have shown the satiating potential of soup. Consuming a preload of low-energy-dense soup, in a variety of forms, has the potential to reduce energy intake in adults.27 Studies have found that liquids in the form of soup can be more satiating than solid foods.28 Several studies have found that consuming soup as a preload can decrease hunger, increase fullness and reduce subsequent test meal intake.2, 28, 29 In Himaya et al.,2 both chunky and smooth soups reduced energy intake at lunch, with the chunky soup having the most effect. In the current study, the smooth soup had the greatest satiating effect, which may be because the rice particles were smaller than the large vegetable particles used in Himaya et al.2 These larger, vegetable particles would retain more liquid and would be expected to remain in the stomach for longer, providing an enduring gastric distension.
Soups have been shown to delay GE as described by Santangelo et al.30 and more recently by Marciani et al.31 Where liquid GE is immediate and exponential, solid food GE is characterized by an initial lag phase followed by a linear emptying phase.32, 33, 34, 35 The lag phase before the food entering the duodenum reflects both redistribution of food from the fundus to antrum and the time it takes for the antrum to reduce solid food to small particles.32 Separation of solids and liquids within the stomach allows faster gastric emptying of the liquid phase compared with the solid, a phenomenon known as sieving. Combining the solid and liquid phases into one thicker consistency such as soup abolishes sieving and delays emptying of the bulk of the meal.31 This increases gastric distension signalling fullness as highlighted in the VAS results.
An interesting outcome of the current study is the increase in GR associated with the soup meal. It is thought that the slower the GE the more reduced the GR is, given that GR is reliant on the delivery of nutrient into the duodenum.36 However, it appears that the greater particulation and hence availability of carbohydrate for absorption overrides any delay in GE so that GR is increased. Our own previous work has indicated that the smaller the particle size of a food, the greater the GR response because of the ease of digestion and absorption.37 Previous research on GR and satiety has indicated that the lower the GR, the more satiating the food,12 yet with soups this appears to have the opposite effect. This may be because of the increased availability of nutrients in the duodenum that can cause the release of satiety hormones.
Insulin is well known to increase satiety.38 An increase in blood glucose as demonstrated in the current study would stimulate insulin secretion, promoting uptake of glucose by muscle and adipose tissue. Other gut hormones including glucagon-like peptide-1 and cholecystokinin are also released rapidly following the arrival of nutrients, especially partially digested nutrients.39, 40 The increased surface area in the smooth soup would have allowed for a much more rapid digestion. Unfortunately, these hormones were not measured in the current study and this does provide scope for future work. Although the increase in satiety is obviously a positive outcome associated with soup ingestion, the increase in GR is not. A high GR to a food results in a high blood glucose concentration and a high insulin-to-glucagon ratio, followed by hypoglycemia, counterregulatory hormone secretion and elevated plasma free fatty acid concentration. This can increase the risk of promoting dyslipidemia, inflammation and endothelial dysfunction.41, 42
It has long been debated whether slowed GE and increased satiety through gastric distension has a greater effect on satiety than the delivery of nutrients, stimulation of nutrient receptors in the duodenum and the release of satiety hormones.9, 43 The current data appear to indicate that soup has the advantage of doing both. It increases satiety as a result of slower GE due to an increased viscosity; and increased GR due to particulation of the food constituents making the nutrients readily available once they do reach the small intestine. In conclusion, soups are satiating because of a combination of delayed GE and rapid availability of nutrients simultaneously.
About this article
European Journal of Nutrition (2014)