Survey
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
* Your assessment is very important for improving the workof artificial intelligence, which forms the content of this project
Calorie restriction wikipedia , lookup
Obesity and the environment wikipedia , lookup
Saturated fat and cardiovascular disease wikipedia , lookup
Human nutrition wikipedia , lookup
Hunger in the United States wikipedia , lookup
Gastric bypass surgery wikipedia , lookup
Table of content Abstract 1. Introduction ............................................................................................................................................. 1 1.1. Research Question......................................................................................................................................... 2 1.2. Clarification of keywords in the research question .......................................................................... 2 1.3. Delimitation of the assignment ................................................................................................................ 2 1.4. Structure of the assignment ....................................................................................................................... 3 1.5. Work method .................................................................................................................................................. 3 1.5.1. Inclusion & exclusion criteria ............................................................................................................................. 6 1.5.2. Selection of papers .................................................................................................................................................. 6 2. Meta-theory plan .................................................................................................................................... 7 3. Theory ........................................................................................................................................................ 8 3.1. Energy intake .................................................................................................................................................. 8 3.2. Macronutrients ............................................................................................................................................... 9 3.2.1. Dietary fat ................................................................................................................................................................... 9 3.2.1.a. Digestion & absorption .................................................................................................................................................... 10 3.2.2. Dietary protein ...................................................................................................................................................... 10 3.2.2.a. Digestion & absorption .................................................................................................................................................... 11 3.2.3. Dietary carbohydrate .......................................................................................................................................... 11 3.2.3.a. Digestion & absorption .................................................................................................................................................... 11 3.2.4. Dietary fiber ............................................................................................................................................................ 12 3.2.4.a. Digestion................................................................................................................................................................................. 12 3.3. Gastric emptying .......................................................................................................................................... 13 3.4. Thermic effect of food ................................................................................................................................ 14 3.5. Satiety .............................................................................................................................................................. 14 3.5.1. The biological process of satiety .................................................................................................................... 15 3.6. Scales ................................................................................................................................................................ 16 3.6.1. Visual analogue scales ........................................................................................................................................ 16 3.6.2. Seven-point bipolar equilateral rating scales ........................................................................................... 17 4. Results ..................................................................................................................................................... 17 4.1 Methodology ................................................................................................................................................... 18 4.1.1. Study design and procedure ............................................................................................................................ 18 4.2. Matrix ............................................................................................................................................................... 19 4.3. Satiety .............................................................................................................................................................. 24 4.3.1. Hunger ....................................................................................................................................................................... 24 4.3.2. Fullness ..................................................................................................................................................................... 25 4.3.3. Subsequent Meal Intake ..................................................................................................................................... 25 4.4. Other parameters ........................................................................................................................................ 26 4.4.1. Desire to eat ............................................................................................................................................................ 26 4.4.2. Pleasantness ........................................................................................................................................................... 27 4.4.3. Prospective consumption .................................................................................................................................. 27 4.4.4. Gastric emptying ................................................................................................................................................... 27 5. Discussion .............................................................................................................................................. 27 5.1 Literature considerations .......................................................................................................................... 41 6. Conclusion ............................................................................................................................................. 44 7. Perspectives .......................................................................................................................................... 44 8. Practice considerations .................................................................................................................... 45 9. List of reference ................................................................................................................................... 47 Abstract Title: Dietary macronutrients as determinants of short-term satiety in adults: a literature review Authors: Marianne Hansen and Sammy Elshikh Background: The prevalence of overweight and obesity has lead to an increased number of noncommunicable diseases; this has sharpened the focus on developing useable tools for weight control. Objective: This thesis aims to provide a review investigating dietary macronutrients effect upon satiety and whether or not it may influence subsequent meal intake. Method: The systematic literature search included 1586 titles. The studies concerned healthy adults and Caucasian subjects. From the 1586 titles, 10 papers where deemed as qualified and included. Results/Conclusion: There was suggestive evidence for protein contributing to a superior effect on satiety compared to the other macronutrients. In addition fat seemed to be the least satiating macronutrient, risking overconsumption of food. The satiating effect of macronutrients does, however seem to be limited to a short time period (3-4 hours). Furthermore it must be investigated whether or not the quantity, meal type, food structure etc. may be of superior influence in regards to satiety. Moreover the thesis has examined how the results may be used for clinical practice, within the profession of health. 1. Introduction Overweight and obesity is becoming an increasing health problem and is affecting a vast number of people around the world. Nearly half of the world's population have become overweight: "In 2014, 39% of adults aged 18 years and older (38% of men and 40% of women) were overweight" (WHO, 2014, p. 79). Furthermore the WHO has classified half a billion people as obese. Due to overweight and obesity, non-communicable and lifestyle diseases has been on the rise causing an estimated 3.4 million deaths per year (WHO, 2014, p. xiv). The reasons for the increased incidence of obesity and overweight have been investigated by numerous health professionals and the results tend to lean toward lifestyle-habits and environmental factors as being the primary cause (Aljuraiban et al., 2015). Research shows that the modern diet has an increased share of fast-foods, which is caloric dense and contains few fibers, fruits, vegetables and has an overall lack of wholesome and healthy food sources (Byrd-Bredbenner, Moe, Beshgetoor, & Berning, 2012, p. 72). This change in nutrition and food sources seems to have caused the increased prevalence in overweight, creating stress on our healthcare system, due to the difficult treatment of obesity. Current methods for effective treatment of obesity, relies on either surgery or low calorie diets (Byrd-Bredbenner et al., 2012, pp. 342–343). By learning and understanding how we most effectively can control weight, we might be able to create safer and better alternatives to the current treatments. The effect of the diet's macronutrient composition on weight maintenance has been investigated as an alternative for effective weight control, within the adult population. However no conclusive evidence for living on a macronutrient specific diet was found, when considering weight maintenance (Fogelholm, Anderssen, Gunnarsdottir, & Lahti-Koski, 2012). The systematic literature review by Fogelholm et al. (2012) showed strong indications for evidence on energy balance and caloric intake being the primary role in weight change and that the macronutrient composition of the diet was ineffective in predicting weight gain. The review by Fogelholm et al. (2012) did point toward other mechanisms that might be relevant in regards to weight control, for example adherence to the specific diet. In studies investigating the macronutrient composition the majority of the diet trails had a relative high dropout rate, which could indicate that some diets are more demanding to follow than others. The results of the studies by Layman et al. (2009), Larsen et al. (2010) and Aller et al. (2014) showed a tendency toward a high protein diet resulting in better adherence. This could mean that even though a specific macronutrient distribution might not be effective in regards to short-term weight control, it might still be favorable for long-term adherence. 1 As adherence to the diet is of uttermost importance, when favoring long-term weight control, this would be relevant to investigate. In certain cases adherence may be related to satiety; the less hunger you feel, the easier it is to stay on track (Byrd-Bredbenner et al., 2012, pp. 332–333). The aim of this thesis is to investigate the satiety rating of the different macronutrients and how they may affect the satiety response in adults. We will hereby examine whether or not a specific macronutrient composition may be correlated to better satiety, thereby becoming a valuable tool when designing adherable weight loss strategies. 1.1. Research Question How does the different macronutrients affect the short-term satiety in healthy adults? 1.2. Clarification of keywords in the research question Macronutrients are categorized into protein, fat, carbohydrate and fiber. Normally fiber is not defined as an individual macronutrient, but is simply separated in this thesis due to its difference in calorie content and digestive process. All the macronutrients are viewed as dietary, meaning they are obtained and present in food. Short-term is defined as ≤24 hours. Satiety is defined according to Gerstein, Woodward-Lopez, Evans, Kelsey, & Drewnowski (2004), as the fullness experienced intra- and inter-meal. Healthy is defined as people without any non-communicable diseases and a BMI ≤30 kg/m2. Adults are defined as people ≥18 years of age. 1.3. Delimitation of the assignment Delimitation was carried out in order to make a clear focus for the thesis and provide a qualitative discussion. This thesis is limited to cover “the macronutrients short-term effect on satiety”; 2 relevant themes, such as the diet's effect on health will therefore not be included due to limited space/characters. We have also chosen to withdraw the link between satiety and hormonal responses, in the results, as it is a complex subject, which would require a study for itself. Furthermore other components such as blood parameters will not be included within this thesis. Neither will we be evaluating the macronutrients within their own subgroups; for example saturated and unsaturated fatty acids, this is also the case for alcohol, which will not be reflected upon, due to limited space. Moreover this thesis will only include human studies, as animal studies measuring satiety may not be directly related to humans, due to the clear physiological differences and means of measurements. In addition children and adolescents below the age of 18 are disqualified from this thesis. Furthermore, people with non-communicable diseases, obesity and smokers are excluded from the included literature, given the fact that it influences several physiological processes (Gregersen et al., 2011). In order to make sure all relevant studies was included in the search; the literature search did not have any restrictions on study types. However this thesis only includes Randomized clinical trials, due to the high ranking in the hierarchy of research evidence (Kjøller, Juel, & Kamper-Jørgen, 2007, p. 371). Studies designs with within-subject, counterbalanced or crossover design were included in the thesis. 1.4. Structure of the assignment You will be able to view our work method and how the included literature was discovered. In order to have a comprehensible understanding of our theme, a section with relevant theory is included. Here you will be able to read about our meta-theory, macronutrients, biological processes and satiety. The results, from the included literature, will be presented in a matrix in order to make it easy to access and examine, in addition the results will be summarized. Moreover the results will be discussed in the following section. This will be the primary part of this thesis, where the content will reflect upon the research question. The thesis will lastly be summed up in the conclusion, thereby providing a concrete answer to our research question. Further perspectives and the professional practice considerations can be viewed at the end of the thesis. 1.5. Work method The work method will be constructed to discover studies regarding satiety and how it is affected by different macronutrients, providing knowledge to weight control and treatment of overweight and 3 obesity. The literature search was conducted in the PubMed and Embase databases. These databases are used, as PubMed is a medical database consisting of a variety of articles within health. A search in Embase was done afterwards in order to secure that relevant European literature also was included. Lastly the literature lists of the studies were looked through, to ensure that we included other significant and relevant studies in the thesis. The search terms used in PubMed are shown in table 1. MeSH terms are used for all the macronutrients in theme 1, in order to narrow down the search to relevant articles concerning dietary macronutrients. In theme 2 the MeSH term ‘satiation’ is used in order to retrieve as many relevant articles as possible. PubMed defines satiation as “Full gratification of a need or desire followed by a state of relative insensitivity to that particular need or desire.” (NCBI, n.d.-b). Furthermore a subgroup for satiation is ‘satiety response’; thereby the search also included articles concerning satiety. Lastly ‘satiety’ was used, as an all field term, to make sure articles without the MeSH term tags ‘satiation’ and ‘satiety response’ also were included. In theme 3 the MeSH term ‘hunger/physiology’ is used, because only articles including the physiological aspect of hunger is relevant for the thesis. Furthermore hunger has its own theme, as we are not interested in articles, which does not include both satiety and hunger. The MeSH term Adult is not included in the PubMed Search, because it is defined as the age of ≥19 (NCBI, n.d.-a). The search terms used in Embase are shown in table 2. In theme 1 the search of dietary macronutrients is made with a broad search, this is done as Embase do not have any Emtrees for the dietary macronutrients. A broad search was chosen instead, because it includes both the words as Emtree and without, e.g. the search term dietary fiber includes: (Dietary AND (Fiber/exp OR fiber)). Hereby it makes sure that we include as many relevant articles as possible in our search. Theme 2, satiation is a broad search, due to the same reasons as in theme 1. Furthermore satiety is included as an Emtree in order to retrieve more relevant articles. In theme 3, hunger is a broad search and not an Emtree, due to the fact that the Emtrees branches for hunger did not cover the areas we were interested in. Lastly theme 4 includes adults as delimitation, since this Emtree is defined as the age of ≥18. 4 Table 1 - Search terms PubMed Theme 1 AND Theme 2 Dietary Proteins (MeSH) Satiation (MeSH) OR OR Dietary Fats (MeSH) Satiety (All Fields) AND Theme 3 Hunger/physiology (MeSH) OR Dietary Fiber (MeSH) OR Dietary Carbohydrates (MeSH) Tabel 2 - Search terms Embase Theme 1 AND Theme 2 Dietary fiber (Broad search) Satiation (Broad search) OR OR Dietary carbohydrate (Broad search) Satiety (Emtree) AND Theme 3 Hunger (Broad search) AND Theme 4 Adult (Lim) OR Dietary protein (Broad search) OR Dietary fat (Broad search) 5 1.5.1. Inclusion & exclusion criteria The inclusion and exclusion criteria of the literature was as following: Participants: Diet: Age ≥18 Fixed test meals (kcal or mass) Healthy Preloads allowed Caucasian Diets regarding Exclusion: Non-communicable diseases Smokers Obese (BMI ≥30) Eating disorders Study type: o Dietary fat o Dietary protein o Dietary Carbohydrate o Dietary fiber Exclusion: Weight Loss programs Sensory-specific satiety Protocol Randomized Control Trial Fast ≥8 or overnight Crossover design Short-term (≤24 hours per session) Within-subject design Counterbalanced 1.5.2. Selection of papers The search terms, for the selection of papers, resulted in 1286 hits, where an additional 300 titles were added from reference lists. The study titles were deemed relevant or unqualified, which resulted in 82 relevant abstracts. The abstracts were read closely and those who did not meet the inclusion criteria were excluded, resulting in a total of 36 papers for skim reading. 18 papers were deemed qualified and were thoroughly read. Eight papers were excluded for not fulfilling the inclusion criteria, leaving 10 papers for this thesis. The exclusion process is shown in figure 1. 6 Figure 1 - Flow-chart of the exclusion process PubMed Embase References Titles (n = 972) Titles (n = 314) Titles (n = 300 ) Total titles screened (n = 1586) 1504 excluded as not being qualified 82 abstracts 46 papers excluded as not being qualified 36 skim-read papers 18 papers excluded as not being qualified 18 full papers 8 papers excluded as not being qualified 10 papers included 2. Meta-theory plan The method that forms this thesis is stated as a literature review and is considered for providing a strong scientific basis in the hierarchy of evidence, which makes it applicable for practical and clinical recommendations (Evans, 2003). The meta-theory and philosophy of science, which constitutes the primary work of this method, is positivism. Positivism is based on the classical empiricism and is described as science being derived through gathering of empirical and 7 observational data. This leads to the concept of cumulative science; within this notion, science is in constant development through the accumulation of knowledge, which is gathered from observations and experiments (Holm, 2013, p. 30). Cumulative science is the very foundation of a literature review, as new knowledge is constantly gathered through new studies/experiments and continues building on the current literature of evidence. Through literature reviews, which examine the current available information, we are able to observe the data and interpret it through reason and logic. This makes us able to put forward theories that may be verified; because only verifiable theories are scientific and can become valid knowledge (Holm, 2013, p. 31). With this approach observations and logic are interconnected, this perspective is the very groundwork of the provided science and an example of how rationalism can be incorporated to structure the observed data, in order to make data processing possible, which is essential for evidence based recommendations (Hjørland, 2011). 3. Theory When investigating satiety, it is necessary to include some background of the factors and theory that are related to the subject. Satiety is a complex matter as there are many aspects to examine, both in regards to physiological functions such as hormones (Byrd-Bredbenner et al., 2012, p. 321), as well as psychological features such as perception and subjective ratings (Gregersen et al., 2011). By exploring current research and evidence related to macronutrients and their satiating effect, we will be including the linked factors that may have an important part in the functions of satiety. 3.1. Energy intake The body can obtain energy from four different nutrient sources: carbohydrates, fats, protein and alcohol. These macronutrients are able to participate in bodily functions, such as building new components and aid in muscular movements (Byrd-Bredbenner et al., 2012, p. 10). The energy provided by the macronutrients is commonly termed as kilocalories (kcal), which will also be the general term used throughout this thesis. The energy contained in a specific food can be measured through practical tools or simply by adding the fuel values from the grams of carbohydrate, fat, protein and alcohol (Geissler & Powers, 2005, pp. 27–28). 8 Fat 9 kcal/g (Ibid) Protein 4 kcal/g (Ibid) Carbohydrate 4 kcal/g (Ibid) Fiber 2 kcal/g (Byrd-Bredbenner et al., 2012, p. 157) Alcohol 7 kcal/g (Geissler & Powers, 2005, pp. 27–28) When considering energy intake and calories in relation to satiety, we must also consider energy density as a factor. Energy density is determined by measuring the foods calorie content and comparing it with the given weight of a food. As fat has the highest calorie content per gram, it is often present in high energy-dense foods, such as fatty meats, crèmes and fried foods. These foods weigh relatively little compared to their calorie content. On the other hand food sources containing low fat, high fiber and high protein are often relatively low in energy density, because of a higher water content, which increases weight, but has no effect on calorie content, due to water being calorie free. Examples of low energy density foods are fruits, greens and low fat meats. As these low energy dense foods can produce greater volume, relative to their calorie amount, they may achieve greater satiety compared to their counterpart, when accounting for energy intake (ByrdBredbenner et al., 2012, p. 52). 3.2. Macronutrients To achieve a greater understanding on the possible satiating effect of the different macronutrients, they will be described through their digestion and absorption. 3.2.1. Dietary fat Fats, also known as lipids, have several features: for example its creamy texture and its ability to add taste to foods, which makes it a popular part of our daily diet. It is present in the majority of the modern society's food choices and it is found in everything from vegetable to animal products. Other than providing food features that can increase the sensation of taste, it also adds important vitamins. Fats have many roles in our body and are essential for different mechanisms to function (Geissler & Powers, 2005, p. 97). It is also the most caloric dense energy source that we can absorb (Geissler & Powers, 2005, pp. 27–28). 9 3.2.1.a. Digestion & absorption The digestion of fat begins when the food enters our mouth; here the enzyme lingual lipase is secreted to begin the digestion of the triglycerides, containing short and medium chain fatty acids. When the fat reaches the stomach, gastric lipase breaks triglycerides into diglycerides, monoglycerides and free fatty acids. However due to fats insolubility, a limited amount is digested in the mouth or stomach. (Geissler & Powers, 2005, pp. 106–107). As the fat moves to the small intestine, the hormone cholecystokinin (CCK) is released from the intestinal cells; this will stimulate the gallbladder to release bile and the pancreas to release lipase and colipase. Bile will then emulsify the fat into small fat droplets, known as micelles, this increases the surface area of the fat. With a larger surface area lipase is able to break down the triglycerides into monoglycerides and free fatty acids; this process is supported by the colipase, which helps the lipase attach to the droplets (Ibid). Phospholipids and cholesterol is also digested in the small intestine: With help of enzymes from the pancreas and the small intestine, phospholipids and cholesterol are broken down to glycerol, fatty acids and other components (Ibid). After the digestion, the fat is absorbed by the brush border of the absorptive cells, in the duodenum and jejunum within the small intestine. Depending on the carbon chain length of the fatty acids and monoglycerides, they will be absorbed by either the lymphatic or the cardiovascular system (ByrdBredbenner et al., 2012, pp. 204–205). After absorption they are reformed into triglycerides; the short and medium chained fatty acids will enter the bloodstream via the cardiovascular system. The long chained fatty acids will enter the lymphatic circulation together with cholesterol and vitamins. Approximately 95% of dietary fat is absorbed; the last 5% is excreted through the large intestine (Ibid). The fat is then transported as different lipoproteins, which enables them to circulate in the blood (Byrd-Bredbenner et al., 2012, p. 206). 3.2.2. Dietary protein Proteins are an essential structure of our body and are found in all of our cells and components: muscle, organs, DNA, blood, enzymes and many others are build on protein, it is therefore an important component within many functions (Byrd-Bredbenner et al., 2012, pp. 339–343). 10 3.2.2.a. Digestion & absorption The digestion of proteins begins in the stomach, where hydrochloric acid is secreted; this acid helps to denature the proteins. The hormone gastrin releases the enzyme pepsin in the stomach, which will begin to break the long polypeptide chains, into shorter amino acid chains through hydrolysis. The partially digested protein then moves to the small intestine. In the small intestine the hormones secretin and CCK is released, these will stimulate the pancreas to release the enzymes trypsin, chymotrypsin and carboxypeptidase. These enzymes will finally break the polypeptides into short peptides and amino acids, which will be absorbed in the small intestine (Geissler & Powers, 2005, p. 80). The absorbed amino acids will be transported to the liver through the portal vein, here they will be used for protein synthesis, energy, or partake in other functions (Byrd-Bredbenner et al., 2012, p. 238). 3.2.3. Dietary carbohydrate Carbohydrates are important for our cells as they provide an easily available energy source and can be used by our muscles and even our nervous system. Carbohydrates can, in small amounts, be stored in blood, muscle and liver, which help to provide a stable blood glucose level and readily available energy (Byrd-Bredbenner et al., 2012, pp. 168–170). 3.2.3.a. Digestion & absorption The carbohydrates are digested in different stages, depending on the structure and whether it is a simple or complex carbohydrate (Byrd-Bredbenner et al., 2012, p. 170). The first digestion process begins in the mouth, where the enzyme salivary amylase breaks down amylose into smaller polysaccharides and disaccharides (Astrup, Bügel, dyerberg, & Sten, 2015, p. 126). When the food reaches the stomach, the salivary amylase is inactivated, by the stomach acidity and no further digestion occurs until the food passes into the small intestine. The pancreas secretes the enzymes pancreatic amylase and dextrinase into the duodenum and the digestion of the amylose continues. Dextrinase also begins the breakdown of amylopectin (Byrd-Bredbenner et al., 2012, pp. 170–171). When the polysaccharides are broken down to disaccharides three enzymes in the wall of the small intestine continuous the digestion. The three enzymes in the small intestine are: (1) lactase, which break down lactose into glucose and galactose, (2) sucrase, which digest sucrose into glucose and fructose and (3) maltase breaking down maltose into two glucose molecules (Ibid). The monosaccharides from dietary carbohydrates do not require further digestion in order to be 11 absorbed. The indigestible carbohydrates continue into the large intestine and will be covered in the section concerning fibers. Fructose is absorbed through a facilitated diffusion, where a carrier protein is used to move the fructose down a concentration gradient. Concentration gradient occurs when the difference of the concentration inside and outside the cell, forces the nutrient into the absorptive cell (ByrdBredbenner et al., 2012, p. 132). Glucose and galactose are absorbed by an active absorption; an active absorption involves a carrier protein and ATP in order to move glucose and galactose, against the concentration gradient and into the absorptive cells (Byrd-Bredbenner et al., 2012, p. 132). The nutrients can be transported against the concentration gradient due to the sodiumpotassium pump, which uses ATP to transport Na+ and K+ ions through the membrane from an area of low concentration into an area with high concentration (Byrd-Bredbenner et al., 2012, p. 488). This makes it possible for glucose and galactose to be pumped into the absorptive cells along with sodium. When the monosaccharide’s enters the intestinal cells they are transported into the liver through the portal vein. In the liver fructose and galactose are converted into glucose. The glucose is then transported to the cells, for energy use, through the bloodstream. The remaining glucose is stored as glycogen in the liver and muscles (Byrd-Bredbenner et al., 2012, p. 172). 3.2.4. Dietary fiber Fibers are composed of non-starch polysaccharides, which enables them different properties and functions (Byrd-Bredbenner et al., 2012, p. 157). Not only do they differ in structure from simple carbohydrates, but they also differ in their digestion process. 3.2.4.a. Digestion The digestion of fibers begins in the large intestine because of the β-bonds, which cannot be broken down by the human enzymes. When the indigestible polysaccharides reaches the large intestine, they are fermented by bacteria into short-chain fatty acids, gas or excreted into fecal waste (ByrdBredbenner et al., 2012, p. 171). The short-chain fatty acids are mainly fermented by soluble fibers and produce fuel for the cells, in the large intestine, and enhance intestinal health (Byrd-Bredbenner et al., 2012, p. 157). 12 3.3. Gastric emptying The rate of gastric emptying is determined by several factors, such as the amount of food, the nutrients, the structure of the food and hormones (Geissler & Powers, 2005, p. 56). Looking at the structure of the food liquids empty more rapidly, from the stomach, than solid The rate the liquid is emptied is in constant proportion with the liquid volume in the stomach, meaning that the rate of 400 mL is twice the rate of 200 mL liquid foods (Shils, Shike, Ross, Caballero, & Cousins, 2006, p. 1181). Furthermore the rate is influenced by: the tonic contractions, the pyloric sphincter resistance and the duodenums capacity to receive liquid. The rate of gastric emptying is further influenced if the liquid contains nutrients, due to feedback from the small intestine: thus the caloric concentration, the pH-value, osmolality and fat content affect the rate (Ibid). High energy density liquids empty in a lower rate than liquids with fewer calories, as the approximate rate for nutrients entering the small intestine is 200 kcal per hour (Ibid). A natural liquid empty faster than an acid liquid, due to the pH difference in the stomach and small intestine. Acid liquids are slowed in order to prevent a change in the small intestine pH-value and thereby maintaining the environment for the pancreatic enzymes (Ibid). Carbohydrates and amino acids regulate the nutrient delivery, through their effect on the small intestine osmolality. Liquids with high osmolality empty slower than liquids with low osmolality, meaning an l00 mL liquid with 0.1 M empties twice as fast as a 100 mL liquid with 0.2 M (Ibid). Liquid containing fat is emptied more slowly compared to aqueous liquid, due to fat binding to solid food particles, thus floating on the surface of aqueous gastric secretion (Ibid). Furthermore lipids inhibits gastric motility, as the hormone signals appearing, when nutrients enters other parts of the gut (Geissler & Powers, 2005, p. 56). Luminal lipids stimulate the release of CCK, peptide YY (PYY), glucagon-like peptides (GLP) and neurotensin. CCK has local effects on the gastric motility, as it stimulates the intestines afferent nerve pathways, resulting in inhibition of gastric activity. PYY is released when short-chain fatty acids are sensed in the ileum, which inhibit gastric motility (Ibid). The gastric emptying of digestible solids occurs at a slower rate than liquids. Before the solids can pass the pyloric sphincter, the food particles diameter needs to be reduced into a maximum of 1-2 mm. This process occurs in the lag phase; here the particles are grinded and mixed. The smaller the particles are the shorter is the lag phase; it can continue up to an hour after the meal. After the lag phase the particles are emptied in a linear rate and in contrast to liquids the rate is independent from the volume in the stomach. Solids larger than 2 mm after the lag phase are emptied from the stomach by the stomach's active contractions (Shils et al., 2006, p. 1181). 13 3.4. Thermic effect of food Macronutrients have different thermogenic values, due to their different ways of digestion, absorption, transportation, storage and metabolization. The thermic effect of food is the amount of energy used to handle the food we consume also known as TEF. This value can account for as much as 10% of our total energy intake, used to digest and absorb etc. This value may of course be different within individuals and even for different meal compositions, as the macronutrients demand different needs of energy for metabolization (Byrd-Bredbenner et al., 2012, p. 317): TEF for fat 0-3% (Ibid) TEF for protein - 20-30% (Ibid) TEF for carbohydrates 5-10% (Ibid) These values are simply due to the structure and functions of the macronutrients, for example it is more energy demanding to metabolize amino acids into fat, compared to transferring absorbed fat into the body's fat storage (Ibid). 3.5. Satiety Satiety is influenced by a multitude of factors; therefore six general concepts need to be clarified, in order to get a better understanding of the biological and physiological processes that affect food consumption. The first concept is Hunger, which is the biological and physiological drive that indicates that the body needs food. It is the organs, hormones, metabolic influences and the nervous system that mainly controls hunger (Byrd-Bredbenner et al., 2012, p. 321). The second concept that needs a calcification is satiation, which is the feeling of fullness developed during a meal, therefore satiation tend to bring the eating episode to an end. Satiation is also referred to as intrameal satiety, because it both reduces hunger and the energy consumed during the meal (Gerstein et al., 2004). The third concept is satiety, which is the fullness developed after the eating episode. Satiety is also referred to as intermeal satiety, because it delays the resumption of eating. Satiety is also commonly referred to as both intra- and inter-meal (Ibid), which is the definition used in this thesis. The fourth concept is appetite. Appetite is the psychological influences that drive the human to eat. It is primarily affected by external factors, such as social influence, emotions, preferences or aversions and environmental factors (Byrd-Bredbenner et al., 2012, p. 321). The fifth concept is sensory 14 specific, which is the change in hedonic response to the sensory properties of the food consumed (Gerstein et al., 2004). Lastly the concept satiety index score (SI%) is the value of subjects satiety ratings based on the food consumptions (Ibid). The relationship between the first four concepts can be explained through feelings, perceptions, observed behavior and operations. When feeling hunger the urge to eat begins, which lead to the observed behavior of eating a snack or a meal. This operation results in the initiation of an eating episode (Blundell & Burley, 1987). Satiation then weakens hunger, which results in the eating coming to an end. This operation is referred to as the termination of the eating episode (Ibid). Appetite begins with the perceived pleasantness of the food and the evaluation of the taste, which leads to the observed behavior of selecting food, change of the eating rate and results in the maintenance of eating (Ibid). Lastly satiety is when there is no wish to eat, which lead to the observed behavior of resistance to further eating. This operation results in inhibition of eating (Ibid). The feelings, perceptions, observed behavior and operations could explain why hunger and appetite do not always correspond to the feelings of satiation and satiety. In some situations a dish is served that evokes the appetite and the eating is maintained or begins even though the feeling of fullness already has occurred. Other times when feeling hungry the food served do not evoke the appetite and the eating does not begin (Byrd-Bredbenner et al., 2012, p. 321). 3.5.1. The biological process of satiety The process of satiety is mainly regulated through the hypothalamus. The hypothalamus is affected by internal signals from different hormones and transmitters, which among others affect satiety (Byrd-Bredbenner et al., 2012, p. 322). The first factor, which elicits satiety, is the sensory aspect of the food eaten, such as smell, flavor and portion size, followed by the knowledge of the food that has been eaten (Ibid). Secondly the stomach and intestines expand; low-energy dense foods, with a high content of water and fiber, causes a greater expansion compared to high-energy foods, such as snacks and oils (Ibid). Lastly the secretion of different hormones during the digestion, absorption and metabolism contribute to the promotion of satiety (Ibid). The hormones influence on satiety is not fully understood, however some gut hormones are known to affect satiety (Murphy & Bloom, 2006). Ghrelin is the only known hormone, which increases appetite, and is therefore often referred to as the hunger hormone. Ghrelin is a peptide hormone that is released from the stomach. It’s levels increases during fasting and decreases after a meal (Ibid). When the levels of ghrelin is 15 regulated it signals to the hypothalamus, resulting in the feeling of hunger during a fasting state or the promotion of satiety after a meal (Byrd-Bredbenner et al., 2012, p. 322). CCK reduces food intake and is secreted from the small intestine after a meal. It seems that cholecystokinin signals, the reduction of food intake, through the vagal nerve (Murphy & Bloom, 2006). PYY and oxyntomodulin (OXM) are secreted from the large intestine after a meal, thus the secretion is reduced during fasting. They reduce food intake and signals to the hypothalamus; furthermore they might signal through the vagal nerve (Ibid). Glucagon-like peptide-1 (GLP-1) is, as PYY and OXM, released through the large intestine and signals to the hypothalamus and possible through the vagal nerve. GLP-1 is released after a meal and reduces hunger; additionally GLP-1 is an incretin, thereby stimulating insulin secretion (Ibid). Pancreatic polypeptide (PP) also reduces hunger and is released from the pancreas after a meal. PP signals to neurons in the brainstem (Ibid). Lastly the hormone leptin has an effect on satiety. It is not a gut hormone, but is produced by the adipose tissue. Leptin signals satiety to the hypothalamus, thus ghrelin works as an antagonist to leptin (Longenbaker, 2014). 3.6. Scales Appetite and satiety are often measured as subjective sensations; this makes the measurements and methods vulnerable to both external and internal influences. Physiological and psychological factors may prove to be of great concern when conducting studies, where reproducibility is of the utmost importance. Here it is central to minimize the factors that could alter the results; this could be done, by excluding restrained eaters or subjects participating in a weight loss. These participants might however be included, if the study would provide a standardized diet prior to the intervention (Flint, Raben, Blundell, & Astrup, 2000). External factors that might influence the results of a given method or test have to be controlled and kept constant, to the best of the abilities, in order to produce valid results. Herein several things must be of concern on test days and the day/days leading up to the test, for example; physical activity, alcohol and food intake. This is essential for reproducibility of the standardized method to be possible (Ibid). 3.6.1. Visual analogue scales Visual analogue scales (VAS) are popularly used in scientific studies to measure subjective sensations, such as appetite. The VAS is set up as a line with a specific length, containing a 16 statement at each end of the line. The statements at each end are set up as opposites or negative and positive statements, for example; I am not full at all vs. I am very full. The test subjects are then to set a mark at their current state, which can be analyzed. As VAS is the main method for satiety ratings in our included literature, it is important that the VAS actually measures the factor, which it is intended to. It is therefore essential that the method is considered valid. In order to inspect the VAS method, the study by Flint et al. (2000) has been examined. In the study by Flint et al. (2000) the VAS was tested on 55 healthy, normal weight men from the age of 19 to 36. The objective was to test the reproducibility and validity of the VAS measurements on appetite sensations, such as hunger and fullness. Flint et al. (2000) found that there was substantial variation in the underlying data that was gathered; this was accounted for as biological and methodological variation. This variation was not deemed as unexpected, as the measurements were concerning subjective feelings. The study concluded that disregarding the mentioned variations, VAS is a valid and reproducible method when concerning appetite ratings and that all appetite parameters seemed to correlate to post energy intake (Ibid). 3.6.2. Seven-point bipolar equilateral rating scales The second rating scale method used was a seven-point bipolar equilateral rating scale (7-point). In this type of scale, one is able to use seven or fewer statements on a bipolar scale, which measures in two directions such as; "not at all" to "extremely”. On the subject of hunger and satiety, an example would be; "not at all hungry" to "extremely hungry". It is important to inquire with a balanced amount of positive and negative statements, in order to reduce response bias and increase validation of the method (Fischer, Colombani, & Wenk, 2004). The 7-point scale has been suggested to create greater retest reproducibility than the VAS, when used on satiety and being easier to understand and use; it was therefore viewed as a compatible method (Holt, Brand Miller, Petocz, & Farmakalidis, 1995). 4. Results In the following section a brief overview of the 10 included studies and their methodology will be presented, followed by a descriptive matrix and a summery of the results. 17 4.1 Methodology All studies recruited their participants in different manners. Some studies included both sexes, while others only included one. All participants were screened before inclusion, in order to exclude factors that could influence the study. Exclusion criteria for all the studies were among others; smokers, obese, restrained eaters and participants with any form of medical condition/illness. This made the included participants more streamlined and made the results comparable across the studies (Gregersen et al., 2011). 4.1.1. Study design and procedure The included studies had comparable designs and procedures. They were all designed as within subject or cross-over designs, were every participant tried each of the test meals; this increased the strength and validity of each study (Flint et al., 2000). All studies had at least one day in between test days, in order to eliminate bias from previous tests. Furthermore all test meals, except the one by Perrigue, Monsivais, & Drewnowski (2009), came in the form of breakfast, followed by an ad libitum lunch/dinner; details of test meals and subsequent meals can be viewed in the matrix. Participants were to cease strenuous exercise prior to the test day and refrain from caffeine and alcohol, some studies included pre-packaged dinners on the evening before the test day - this was however not a criteria for all studies, neither was it seen as a critical factor for allowing comparison (Flint et al., 2000). Before a test day, the subjects were fasted with a minimum of eight hours. At arrival the participants were screened for compliance before presentation of the test meal as well as assessed for hunger and other parameters. Subjective hunger ratings were assessed, by using the VAS Score, except the one study by Fischer, Colombani, & Wenk (2004), who used the 7-point score. Both methods are deemed as valid and reproducible (Holt et al., 1995). The different studies did however have different times for when to conduct measurements and when to serve the different test meals. Some studies also included other parameters than the standard hunger and fullness, if relevant this will be shown in the result summary. In order to make the results from the included studies more comparable to each other, the authors of this thesis have redefined the name of the different test meals. The renaming of the test meals has been done in accordance to the NNR recommendations of macronutrient intake. The definitions are 18 as following: low fat <25E%, high fat >40E%, low protein <10E%, high protein > 20E%, low carbohydrate <45E%, high carbohydrate >60E% and high fiber >3/MJ (Nordic Council of Ministers, 2008, pp. 25–27). Fiber is only defined as high, due to the fixed recommendation by NNR, which makes it difficult to define low/moderate range. Furthermore this thesis defines significance as a p-value of ≤0.05. 4.2. Matrix M: Male HF: High fat W: Women LF: Low fat Age: Years BMI: Kg/m HP: High protein 2 LP: Low protein PDS: Prior diet standardization HC: High carbohydrate F: Fat LC: Low carbohydrate P: Protein HFi: High fiber C: Carbohydrate HR: Hunger rating Fi: Fiber SM: Subsequent meal Q: Quantity (g/mL) *: Data for males/females -: No data +: The study contains two experiments Bold text: marks the superior diet ↑ indicates significant higher rating/intake compared with other test meals ↓ indicates significant lower rating/intake compared with other test meals → indicates no significant difference between the test meals 19 Tabel 3 - Matrix Study Design Clegg & Randomized Shafat , balanced, 2010 Participants Intervention No. Sex Age BMI 9 M 25.5 - PDS Test meal F: 70%, P: 7%, C: 25%, Fi: 1.2 g, Result Meal name HF/LP/LC Q: 222 g, Kcal: 771 HC vs. HF/LP/LC & HF: single-blind F: 29%, P: 11%, C: 65%, Fi: 1.2 g, (p<0.05) study Q: 357 g, Kcal: 789 ± 1.6 HC F: 43%, P: 13%, C: 47%, Fi: 1.2 g, Cotton, Within- Burley, subject Weststra design 16 ------ M ------- HF vs. HC: (p<0.01) HF Dropout: 0% LF/HC HR 3h: 22.87 F: 23%, P: 11%, C: 70% Fi: - 0.35 ± 0.46 Q: 580 g, Kcal: 440 ------------ F: 57%, P: 6%, C: 39%, Fi: - LF/LP/HC vs. HF/LP/LC: HF/LP/LC 12 M (p=0.03) LF/LP/HC vs. LF/HC: Q: 640 g, Kcal: 803 te, & 1994 SM: Fixed & Ad libitum 21.06 ± --------- HR 7hr: SM: Q: 222 g, Kcal: 438 Blundell, HR 3hr: LF/LP/HC LF/HC vs. HF/LP/LC: 23.4 22.87 F: 13%, P: 6%, C: 86%, Fi - ± 0.9 ± 0.46 Q: 735 g, Kcal: 803 HR 10h: SM: Ad libitum SM: -----------------------HR 2h: LF/LP/HC vs. HF/LP/LC & LF/HC: (p<0.01) LF/HC vs. HF/LP/LC: SM: 20 LF/LP/HC vs. HF/LP/LC: (p<0.02) LF/LP/HC vs. LF/HC: (p<0.04) LF/HC vs. HF/LP/LC: Dropout: 0 % 40 M, W 27 ± 5 Dougkas Randomized 24.3 ± 1.6 X F: 63%, P: 9%, C: 28%, Fi: - & , within- *Q: 667/533 g, *Kcal: 502.3/402 Ostman, subjects, 2- F: 42%, P: 40%, C: 19%, Fi: - 2016 level *Q: 667/533 g, *Kcal: 502.3/402 factorial F: 20%, P: 9%, C: 71%, Fi: - design *Q: 667/533 g, *Kcal: 502.3/402 F: 13%, P: 40%, C: 47%, Fi - HF/LP/LC HR 3.5h: SM: HF/HP/LC Dropout: 10% LF/LP/HC LF/HP *Q: 667/533 g, *Kcal: 502.3/402 F: 25%, P: 24%, C: 50%, Fi: - HP *Q: 667/533 g, *Kcal: 502.3/402 SM: Ad libitum Fallaize, Randomized 31 M 21.7 23.1 ± 1.15 ± 2.65 F: 57%, P: 23%, C: 23%, Fi: - Wilson, , 3-way Gray, crossover F: 26%, P: 11%; C: 64%, Fi: - Morgan, design Q: -, Kcal: 331 HF/HP/LC HR 3h: HF/HP/LC vs. HF/LP: Q: -, Kcal: 327 HC (p<0.05) HF/HP/LC vs. HC: HF/LP HF/LP vs. HC: & F: 49%, P: 6%, C: 46%, Fi: - Griffin, Q: -, Kcal: 328 HR 8.5h: 2013 SM: Ad libitum SM: HF/HP/LC vs. HF/LP: (p=0.03) Dropout: 3% 21 17 M 26.5 Fischer Repeated 21.9 ± 1.7 X et al., measures, 2004 counterbalan F: 0%, P: 100%, C: 0%, Fi: - ced cross- Q: 400 mL, Kcal: 400.8 over design F: 100%, P: 0%, C: 0%, Fi: - ± 3.3 20 M, W 24.1 F: 0%, P: 0%, C: 100%, Fi: - LF/LP/HC LF/HP/LC vs. HF/LP/LC: Q: 400 mL, Kcal: 400.8 21.7 ± 2.2 LF/HP/LC (p<0.001) LF/HP/LC vs. LF/LP/HC: HF/LP/LC (p<0.05) 400 mL, Kcal: 400.8 SM: SM: Ad libitum Dropout: 12% Kristens Open- F: 34%, P: 21%, C: 45%, Fi: 3.6 g en et al., labeled 2010 randomized F: 28%, P: 20%, C: 52%, Fi: 11.7 g multiple Q: 181.4 g, Kcal: 480 crossover F: 35%, P: 21%, C: 44%, Fi: 2.2 g design Q: 213.5 g, Kcal: 480 ± 3.8 HR 2.45h: HP SM: Q: 169.9 g, Kcal: 480 F: 31%, P: 21%, C: 48%, Fi: 5 g HR 3h: HFi Dropout: 20% HP HP Q: 227.2 g, Kcal: 480 SM: Ad libitum Perrigue, Within- 40 M, W - - X F: 12%, P: 11%, C: 60%, Fi: 6 g Monsiva subject, 472 mL, Kcal: 440 is, & counterbalan F: 12%, P: 11%, C: 60%, Fi: 0 g Drewno ced design Q: 472 mL, Kcal: 440 LF/HFi HR 2h: LF/HFi, LF, LF/HP/LC/HFi LF & LF/HP/LC vs. LF/LP/HC: (p<0.05) wski, F: 0%, P: 27%, C: 40%, Fi: 6g LF/HP/LC/H SM: 2009 Q: 472 mL, Kcal: 180 Fi Dropout: 5% F: 0%, P: 27%, C: 40%, Fi: 0 g LF/HP/LC Q: 472 mL, Kcal: 180 F: 0%, P: 0%, C: 98%, Fi: 0 g LF/LP/HC Q: 472 mL Kcal: 180 SM: Ad libitum 22 Poppitt, Randomized 12 W - - X F: 19%, P: 11%, C: 68%, Fi: - Mccorm , Within- Q: 712 g, Kcal: 528 ack, & subject F: 69%, P: 11%, C: 21%, Fi: - Buffenst design Q: 732 g, Kcal: 504 ein, F: 19%, P: 59%, C: 20%, Fi: - 1998 Q: 786 g, Kcal: 456 LF/HC HR 2h: SM: HF/LC LF/HP/LC vs. LF/HC & HF/LC: (p<0.05) LF/HP/LC Dropout: 0 % LF/HC HR 3h: SM: Ad libitum Rolls et Within- al., 1991 subjects, 28 M, W M: 26.4 - ± 4.1 F: 5%, P: 14%, C: 81% Fi: - repeated F: 65%, P: 15%, C: 20%, Fi: F: 25.2 measures W Dropout: 0% HP HR 2h: SM: Ad libitum Rolls, Within- Hetherin subject, gton, & counterbalan F: 8%, P: 9%, C: 70%, Fi: - Burley, ced design Q: 390 g, Kcal: 300.3 1988 10 HF/LC *Q: 500/350 g, *Kcal: 510/357 ± 4.4 design SM: *Q: 500/350 g, *Kcal: 510/357 23.1 21.0 ± 4.1 ± 1.13 F: 25%, P: 75%, C -, Fi: Q: 210 g, Kcal: 298.3 F: 97%, P: 5%, C: - Fi: - HP, LF/LP/HC, HF/LP & LP LF/LP/HC LF/LP/HC & HP vs. HF/LP: HF/LP LP/HC Q: 94 g, Kcal: 299.8 F: 38%, P: 5%, C: 60%, Fi: - (p<0.01) LF/LP/HC & HP vs. LP: (p Q: 66 g, Kcal: 289.7 F: -, P: 2%, C: 91%, Fi: vs. LP/HC: (p<0.01) < 0.01) SM: LP LF/LP/HC & HP vs. HF/LP, Q: 68 g, Kcal: 299.9 LP/HC & LP: (p<0.05) SM: Ad libitum (Lunch) Dropout: 0 % 23 4.3. Satiety This section summarizes the results of the test meals effect on hunger, fullness and subsequent meal intake. These measurements will be stated according to their relation to satiety and whether or not they were deemed significant. It must be mentioned that the procedure and questionnaires vary within the different studies. 4.3.1. Hunger All the studies investigated the combination between hunger ratings and different macronutrient intake. Six studies found a significant difference between the macronutrients and the hunger ratings. Moreover the significant difference was only found within a 3-hour period. Clegg & Shafat (2010), found that a high carbohydrate (HC) meal reduces hunger, to a higher extent, than a meal high in fat, low in protein and low in carbohydrates (HF/LP/LC) or than a meal high in fat (HF) (p<0.05). However the differences in hunger was only significant within the first 3 hours. According to Cotton et al. (1994), experiment 1, a meal low in fat and in protein and high in carbohydrate (LF/LP/HC) reduces hunger more than a HF/LP/LC meal (p=0.03). However as in Cleeg’s study, there was only a significant difference, between the test meals, during the first 3 hours. In experiment 2, Cotton found a decrease, within a 2-hours range, in hunger after a LF/LP/HC meal compared to a HF/LP/LC meal and a low fat, high carbohydrate (LF/HC) meal (p<0.01). Fallaize et al (2013) found that a high fat, high protein and low carbohydrate (HF/HP/LC) meal decreases hunger to a higher extent than a meal high in fat and low in protein (HF/LP) (p<0.05); although the significant difference was only valid within a 3-hour range. According to Fischer et al. (2004) a low fat, high protein and low carbohydrate (LF/HP/LC) meal reduces hunger more than a HF/LP/LC meal (p<0.001) and a LF/LP/HC meal (p<0.05). Perrigue et al. (2009) found a lower reduction in hunger after a LF/LP/HC meal compared to a low fat, high fiber (LF/HFi), a low fat (LF), a low fat, high protein, low carbohydrate, high fiber (LF/HP/LC/HFi) and a low fat, high protein, high carbohydrate (LF/HP/HC) meal (p<0.05). Rolls et al. (1988) found a decrease in hunger after a high protein (HP), a high, fat, low protein (HF/LP), a LF/LP/HC and a low protein (LP) meal compared to a low protein, high carbohydrate (LP/HC) meal (p<0.01). Furthermore the LF/LP/HC and HP test meals did result in a higher decrease in hunger compared to a HF/LP (p<0.01). Lastly Rolls et al. (1988) found that a LF/LP/HC and a HP had a higher effect on hunger 24 compared to a LP meal (p<0.01). Dougkas & Ostman (2016), Kristensen et al. (2010), Poppitt et al. (1998) and Rolls et al. (1991) found no significant difference between the compared test meals. 4.3.2. Fullness Nine of the included studies investigated the macronutrients effect on fullness. Six studies found a significant difference between the macronutrients and their effect on fullness. Similar to the findings of hunger, the effect only occurred within a 3-hour range. Clegg & Shafat (2010) showed a similar result to hunger on fullness; HC was rated higher than HF on their 3-hour rating (p<0.05). Cotton et al. (1994) found no effect on fullness from the 10-hour rating, however the 3-hour rating did show a significant difference between the LF/LP/HC and the HF/LP/LC (p=0.03), as LF/LP/HC resulted in increased fullness. Fullness was not measured in experiment 2. Dougkas & Ostman (2016) managed to find a significant difference on the 3,5 hour fullness rating between HF/LP/LC and LF/HP (p=0.03), where LF/HP increased the fullness rating. Fallaize et al. (2013) fullness rating results were different, when comparing HF/HP/LC and HF/LP (p<0.05), as HF/HP/LC gave a greater rating of fullness from the subjects. Perrigue et al. (2009) found that a LF/LP/HC meal resulted in a lower 2-hour rating of fullness compared to a LF/HFi, a LF, a LF/HP/LC/HFi and a LF/HP/LC meal (p<0.05). Furthermore the subject was less full after a LF/HP/LC meal compared to a LF/HFi and a LF/HP/LC/HFi meal (p<0.05). Rolls et al. (1988) compared HP and LF/LP/HC vs. LP/HC and HF/LP, LP and found that HP and LF/LP/HC increased fullness significantly in the 2-hour period (p<0.05). Kristensen et al. (2010), Poppitt et al. (1998) and Rolls et al. (1991) found no difference in the test meals effect on fullness. No data was found in Fischer et al. (2004). 4.3.3. Subsequent Meal Intake The different test meals effect on subsequent meal intake was measured in all the included studies. Four studies found a significant difference between the macronutrients and subsequent meal intake. Similar to the other results the effect only occurred within a 3-4-hour period. Clegg & Shafat (2010) found a significant effect on the subsequent meal intake between the HF test meals, which caused a higher energy intake than the HC test meal (p<0.01). In the study by Cotton et al. (1994) the two experiments provided different results; in experiment 1 there was no difference in the subsequent meal intake between the test meals, however in 25 experiment 2 the LF/LP/HC test meal led to a reduced energy intake compared to the HF/LP/LC (p<0.02). Furthermore the LF/LP/HC test meal also led to a reduced energy intake in comparison to the LF/HC (p<0.04). Fallaize et al. (2013) found a significant effect between two of their test meals; HF/HP/LC proved to result in a lower energy intake in the subsequent meal, when compared to the HF/LP (p=0.03). Poppitt et al. (1998) also showed a significant difference between their LF/HP/LC vs. the LF/HC and high fat, low carbohydrate test meal (HF/LC) (p<0.05), here the LF/HP/LC test meal resulted in a reduced subsequent meal intake at lunch. Rolls et al. (1988) discovered a different effect between the LF/LP/HC and HP vs. the HF/LP, LP/HC and LP (p<0.05), as the LF/LP/HC and HP resulted in a lower subsequent meal intake. Dougkas & Ostman (2016), Fischer et al. (2004), Kristensen et al. (2010), Perrigue et al. (2009) and Rolls et al. (1991) found no difference on the test meals effect on subsequent meal intake. 4.4. Other parameters When assessing satiety other parameters than hunger, fullness and subsequent intake were measured. Seven of the included studies measured on the subject's desire to eat and the pleasantness of the test meals. Six studies measured the prospective consumption and two studies included gastric emptying as a parameter. 4.4.1. Desire to eat Four out of seven studies found a significant difference, of the subjects desire to eat, after the test meals. Two studies found no difference after the test meal and lastly no data was found in the study by Kristensen et al. (2010). Clegg & Shafat (2010) found that a HC meal resulted in a lower desire to eat compared to a HF meal (p<0.05), however there was only found a significant difference within the first 3-hours. Likewise Cotten et al. (1994), experiment 2, showed a lower desire to eat after a LF/HP/HC meal compared to a LF/HC and a HF/LP/LC meal (p<0.05). A LF/HP/LC meal resulted in less desire to eat compared to a HF/LP/LC meal (p<0.01) in the study by Fischer et al (2004). According to Perrigue et al. (2009) a LF/HFi meal result in a lower desire to eat compared to the LF/LP/HC test meal (p<0.05). Dougkas & Ostman (2016) and Rolls et al. (1991) found no difference between the test meals and the subject's desire to eat. 26 4.4.2. Pleasantness There is an overall agreement, within the studies measuring, on the test meals pleasantness. Five out of the seven studies found no difference on the pleasantness of the test meals, these studies includes Cotton et al. (1994), Dougkas & Ostman (2016), Kristensen et al. (2010), Poppitt et al. (1998) and Rolls et al. (1991). No data was found in the studies by Clegg & Shafat (2010) and Fischer et al (2004). 4.4.3. Prospective consumption Three out of the six studies found no difference in prospective consumption. One study found significance and lastly in the studies by Cotton et al. (1994) and Fischer et al. (2004) no data was found. Fallaize et al. (2013), Kristensen et al. (2010) and Rolls et al. (1991) was in agreement and found no difference in the prospective consumption after the test meals. Dougkas & Ostman (2016) on the other hand found that a LF/HP and a HF/HP/LC meal resulted in a lower rating of prospective consumption compared to a LF/LP/HC meal (p<0.05) 4.4.4. Gastric emptying Two studies investigated gastric emptying after the test meals. Clegg & Shafat (2010) found that a HF/LP/LC meal slowed gastric emptying compared to a HF (p<0.05) and a HC (p<0.01) meal. Likewise Fischer et al. (2004) showed a HF/LP/LC meal slowed gastric emptying compared to a LF/HP/LC (p<0.01) and a LF/LP/HC (p<0.05) meal. 5. Discussion This thesis investigated the macronutrients effect on satiety. Looking at the results from the literature, the answer to the research question is inconclusive, however the results suggest that protein might be superior, compared to the other macronutrients. In order to conclude on the results, procedures and limitations will be discussed and investigated to see whether other factors have influenced the results. 27 Fat & Satiety In the study by Clegg & Shafat (2010) two different high fat meals were tested: HF/LP/LC and HF. Both of these were significantly less effective in decreasing hunger when compared to a HC test meal (p<0.05). The measurements on fullness also showed the HF test meal to be less effective when compared to the HC (p<0.05), moreover the HF test meal increased intake at the SM compared to the HC (p<0.01). The results from Clegg & Shafat (2010) would indicate that fat, as a macronutrient, has an inferior effect on satiety compared to carbohydrate. However components of the included test meals were more complex than just macronutrient composition; the test meals also differed in quantity and calories. Within these differentials, quantity and portion size may come in to play, as it can affect the following food intake and satiety. This was shown by Kral (2006), who found a significant correlation between portion size and satiety. The effect of portion size would seem to be present in Clegg & Shafat (2010), as the most satiating of the three test meals, the HC, was also the heaviest with 135g more than the others; this would cause greater gastric distention and may have been the reason for the decreased hunger, increased fullness and decreased SM intake compared to the HF test. The HF/LP/LC and HF test meals did however not differ in quantity, but in calories, which did not seem to significantly influence the results even though the difference was 333 calories. It would therefore appear that not only is fat less satiating than carbohydrate, but also that meal quantity is more effective in influencing satiety than meal energy content. The results from Clegg & Shafat (2010) can be confirmed by Cotton et al. (1994), who came to similar results in their study. In both experiment 1 and 2 the hunger rating of the HF/LP/LC test meal was significantly less effective compared to the LF/LP/HC (p=0.03 & p<0.01) and showed no effect, when HF/LP/LC was compared to the LF/HC. In experiment 1, fullness also decreased in the HF/LP/LC vs. the LF/LP/HC (p=0.03), but the SM intake did however not differ between the test meals. At first it would seem that there might not be a correlation between hunger, fullness and SM intake. However when examining the study procedure, one will notice the longer time interval between the test meals and lunch, compared to experiment 2. The results from experiment 2, could indicate that the satiety abilities of a meal only lasts for a short period of time as the HF/LP/LC had a significantly higher SM intake compared to the LF/LP/HC (p<0.02) in a 2hour period, but did not differ from LF/HC. This outcome confirms the limited suppression fat has on satiety in the short-term, compared to a high carbohydrate meal. Once again it should taken into consideration, that there are relative differences in quantity between the meals; the most satiating was also the heaviest; LF/LP/HC contained 95g more than HF/LP/LC and 155g more than LF/HC. 28 This would once again pose the hypothesis of portion size being the primary factor. This relation is also confirmed by Kral (2006), who examined the effect of portion size on satiety and found a clear correlation between increased portion size leading to increased satiety. Yet HF/LP/LC still showed no effect for better or worse, when compared to LF/HC in either of the parameters, even though HF/LP/LC was 60g heavier and contained 363 calories more than LF/HC; an indication of fats inferior effect on satiety. Poppitt et al. (1998), were not able to show a significant difference on hunger rating or fullness rating between the test meals, but did manage to reach significance in SM intake. As to why there was no measurable effect on hunger and fullness is difficult to determine. However the HF/LC was shown to increase SM intake compared to LF/HP/LC (p<0.05). As seen in Clegg & Shafat (2010) and Cotton et al. (1994) the most satiating meal was the heaviest, with 54g between LF/HP/LC and HF/LC. Yet a test meal dominant in fat content showed to be less effective in affecting satiety, than a test meal dominant in protein or carbohydrate, as there was found no significant effect between HF/LC and LF/HC, even though HF/LC was a heavier meal. This difference should according to Kral (2006) have led to higher satiety from the HF/LC, yet this was not the case; a similar result was also seen in Fischer et al. (2004). In the study by Fischer et al. (2004) portion size was not a relevant factor, as the test meals served were of equal volume and calorie content. The results still showed fat as inferior in hunger rating when HF/LP/LC was compared to LF/HP/LC (p<0.001), yet no difference between HF/LP/LC and LF/LP/HC was found. SM meal intake was not significantly different for any of the test meals. The reason for the lack of significant difference in SM intake might very well be due to the >3-hour interval between the test meal and lunch. This duration was most likely to long for the test meals to withhold their short-term difference in satiety; this was also seen in experiment 1, by Cotton et al. (1994). However the result does indicate for fat having a weaker effect on satiety than protein, when served in pure macronutrient forms. Dougkas & Ostman (2016) found a significant effect on the fullness ratings, where LF/HP increased the fullness compared to the HF/LP/LC. As there was accounted for both quantity and calorie content this would suggest a superior effect of protein vs. fat. In contrast Rolls et al. (1991) did not manage to find a significant difference on fullness, which likely is do due to the comparison concerning high fat vs. high carbohydrate. It would indicate that these two macronutrients are relatively alike in their effect on fullness, or at least that their abilities are not significantly different if served in similar quantity and calorie content. Neither study found a 29 difference whatsoever on hunger ratings or SM intake between the test meals, both of these studies used equal weight and iso-caloric test meals, as well as serving them in a liquid form. The results somewhat contradict the previous findings and shows no difference in short-term satiety sensations after consuming fat, protein or carbohydrate-rich test meals. This contradictory result may nonetheless be due to the form the preloads were served in. It has been seen that liquid forms may have a weaker effect on satiety, and withhold an even shorter period of effect due to the quicker gastric emptying of liquids contra solid foods (Shils et al., 2006, p. 1181). This may explain why the studies were not able to find a significant difference. Fallaize et al. (2013) and Rolls et al. (1988) had quite different results on fats satiety capabilities compared to the rest of the included studies. Both studies found high fat test meals to be superior, when compared to some of the other test meals; this result can however be explained by other factors than fats influence on satiety. Fallaize et al. (2013) found HF/HP/LC to decrease hunger compared to HF/LP (p<0.05) and increase fullness compared to HF/LP (p<0.05), HF/HP/LC also decreased SM intake compared to HF/LP (p=0.03). Both of these test meals were iso-caloric and high fat, yet one proved to be more effective at increasing satiety than the other. The outcome may likely be due to the substantial difference in the protein content; 6% in the HF/LP to 23% in the HF/HP/LC. As seen in the previous studies, fat has a weak influence on satiety, but there is suggestive evidence towards protein having a strong influence on satiety, which was also seen in Fischer et al. (2004) and Poppitt et al. (1998). Rolls et al. (1988)’ results are still similar to the other studies suggesting fat as a poor influence on satiety. The HF/LP test meal scored lower on hunger and fullness compared to the LF/LP/HC and HP (p<0.01 & p<0.05). In addition SM intake was also higher following HF/LP compared to LF/LP/HC and HP (p<0.05); this is in agreement with the previous literature, suggesting fat as inferior in affecting satiety. The study did in addition show that all test meals including HF/LP was rated superior in affecting hunger compared to the LP/HC (p<0.01), however these test meals were extremely different even though they were designed as iso-caloric. The weight and food-type of the test meals differ considerably, as the most satiating test meals were almost six and three times heavier than the HF/LP, making portion size a factor; 390g and 210g vs. 66g. Furthermore the food types differed from what would be considered a main meal to snacks; pasta and chicken vs. cream cheese, Turkish delight and confectionary. This might very well influence the subjective rating of the food, as sensory properties may influence satiety (Buckland, James Stubbs, & Finlayson, 2015). With these differences in the test meals, it is difficult to 30 determine the reason behind the results, as it very well may be caused by multiple confounding factors. There seems to be a tendency toward high fat meals being less satiating than other macronutrient composition meals. Clegg & Shafat (2010), Cotton et al. (1994) and Rolls et al. (1988) all showed that high fat meals were inferior when compared to high carbohydrate meals, in addition Poppitt et al. (1998), Fischer et al. (2004), Fallaize et al. (2013) and Rolls et al. (1988) showed that this was also the case when comparing high fat to high protein. Only two studies found no effect on hunger when comparing the different macronutrient compositions; Rolls et al. (1991) and Dougkas & Ostman (2016). In order to explain the outcome it is important to shed light upon what might be the mechanisms resulting in macronutrients different effect on satiety. Protein & Satiety Fallaize, et al. (2013) found the HF/HP/LC test meal to reduce hunger and increase fullness significantly compared to HF/LP (p<0.05 & p<0.05), however no difference between the HF/HP/LC and the HC test meal was found. In addition SM intake was also decreased after the HF/HP/LC compared to HF/LP test meal (p=0.03). All three test meals were iso-caloric with different macronutrient compositions; portion size/quantity was however not mentioned and may have been a factor. Kral (2006) has previously shown the significant effect of portion size correlating to satiety; the lack of mentioning this factor in Fallaize, et al. (2013) is therefore a rather noteworthy limitation. Furthermore test meal type might also have been a confounding factor, as the subjective rating and perception of the test meals might also influence the result; eggs vs. croissant. Buckland et al. (2015) has shown that foods not only have an influence on physiological satiety when eaten, but also on psychological satiety when presented. The very perception of foods ability to satiate may likely influence any results and ratings and could therefore be a limiting factor when presenting different foods. However it is still interesting to see that HF/HP/LC was rated more effective on satiety compared to the HF/LP, as they are both high fat meals, with protein content being the primary difference. Depending on whether or not quantity and meal type was a limiting factor this would pose the hypothesis of protein being more satiating than fat. Fischer et al. (2004) results supported the data by Fallaize et al. (2013), as it showed protein having a superior effect on hunger, in iso-caloric and iso-volumetric test meals. The LF/HP/LC reduced hunger compared to both HF/LP/LC (p<0.001) and LF/LP/HC (p<0.05). The 31 test meals did however not differ in their effect on SM intake, as in Fallaize et al. (2013). This is most likely due to the amount of time between the test meals and lunch, which is longer than 3 hours. The results from Perrigue et al. (2009) only found a significant difference of hunger ratings in LF/LP/HC, which was rated higher, in comparison to the rest of the test meals (p<0.05). This might point toward the satiating effect of protein, as the low protein meal increased hunger more than all the other test meals. There was conversely no difference between LF/HFi, LF, LF/HP/LC/HFi and LF/HP/LC in hunger ratings, even though there were differences in both macronutrient composition and caloric content, none of these were however low in protein. All meals were of same quantity, which would suggest that if protein content is adequate, then other factors will have a limited influence on satiety. Fullness was also differential between the meals; LF/LP/HC decreased fullness compared to LF/HFi, LF, LF/HP/LC/HFi and a LF/HP/LC meal (p<0.05), in addition LF/HP/LC also decreased fullness compared to LF/HFi and a LF/HP/LC/HFi meal (p<0.05). It would seem that fullness is not only influenced by protein, but also fiber. Even though LF/HP/LC was similar to LF/HP/LC/HFi in protein and energy content, it was not as effective to produce fullness - being that fiber content was the only difference: it would seem that fiber also affects satiety to a certain degree. SM intake did not differ significantly between any of the included test meals, which might be due to the time interval between the test meals. Rolls et al. (1988) supports the results from Fallaize et al. (2013), Fischer et al. (2004) and Perrigue et al. (2009) and found high protein meals to be superior, compared to other macronutrient compositions. Hunger ratings resulted in HP decreasing hunger compared to LP/HC (p<0.01), HF/LP (p<0.01) and LP (p<0.01), while also increasing fullness compared to LP/HC and HF/LP and LP (p<0.05). Furthermore HP decreased SM intake compared to HF/LP, LP/HC and LP (p<0.05). These results do show a clear indication of protein being superior in inducing satiety, however the test meals did differ significantly in weight and food-type, making it difficult to determine the actual cause for the results, as these factors have previously been seen to affect satiety (Kral, 2006). Even though Poppitt et al. (1998) did not manage to find significant differences between their test meals on hunger and fullness, they still managed to find high protein as superior in SM intake. LF/HP/LC significantly decreased SM intake compared to both LF/HC & HF/LC (p<0.05). The high protein test meal would seem to be most effective at inducing satiety, but it was also the heaviest meal - making a solid conclusion on the effect of macronutrient composition 32 difficult, as the portion size is of strong influence on satiety and could have been the determining factor creating the significant difference (Kral, 2006). Dougkas & Ostman (2016) showed a difference on fullness, where protein was superior: LF/HP vs. HF/LP/LC (p=0.03), but did however not find a difference on hunger or SM intake between the test meals. The test meals were of equal weight and iso-caloric content, however served as liquid forms, which may or may not have been the cause for a lack of difference in hunger and SM intake. The ambiguous part of the result is why LF/HP were the only test meal rated higher than HF/LP/LC in fullness, as both HF/HP/LC and HP was also high protein meals. HF/HP/LC was however low in carbohydrate and high in fat compared to the LF/HP, suggesting that carbohydrate content is more satiating than fat, in addition HP also contained more fat and less protein than LF/HP, confirming the low satiating power of fat contra both protein and carbohydrate. Kristensen et al. (2010) found no difference between any of the ratings on hunger, SM intake or fullness. The included meals were all of different quantity and iso-caloric, furthermore all but one was high protein. The test meal moderate in protein was still 20% vs. the others containing 21% - making them differ by as little as 1%. In addition both fat and carbohydrate content only differed within a range of 8%, making the meals fairly similar apart from quantity and fiber content. It would therefore seem that if protein content is adequate and the macronutrient distribution is somewhat equal, combined with an iso-caloric energy content, then the satiety parameters would not differ significantly. In this case Protein content seems to mask the effects of quantity and to a certain degree making it an insignificant factor concerning satiety. This is however contradictory to the results from Kral (2006) where portion size was clearly correlated to satiety; yet this result does indicate the complexity of satiety, as a phenomena and how different factors be of greater importance than others. There are strong indications for protein being the most satiating macronutrient, when compared within test meals. Fallaize et al. (2013), Fischer et al. (2004), Perrigue et al. (2009), Rolls et al. (1988), Poppitt et al. (1998) and Dougkas & Ostman (2016) all showed superior effects of protein in either decreasing hunger, increasing fullness or decreasing SM intake compared to fat and carbohydrate. Kristensen et al. (2010) was the only study, which did not find any effect, however this was likely due to similar macronutrient composition of the test meals. 33 Carbohydrates & Satiety In the study by Clegg & Shafat (2010) a HC meal was compared with high fat meals. They found that a HC meal reduced hunger to a higher extent, within a 3-hour period, compared to a HF/LP/LC (p<0.05) and a HF meal (p<0.05). Clegg & Shafat (2010) used three similar breakfasts; however the quantity and the calorie content differed between the three test meals. The higher quantity and calorie content might be the reason for the HC meal’s reduction in hunger. Interestingly the results from hunger are not in agreement with the fullness ratings, where no difference was found between the HC and the HF/LP/LC test meal. This can be explained by the difference in calorie content, indicating that calories might have an effect on fullness. Another reason could be the difference in gastric emptying of the foods, as carbohydrates are emptied at a higher rate. Resembling results were found for the SM intake, indicating that the effect on satiety only exists within a 3-hour period. The results from Clegg & Shafat (2010) suggest that HC meals are more satisfying than HF, however it cannot be overruled that the higher satiety is caused by the difference in quantity. In the study by Cotton et al. (1994) two HC test meals were tested: LF/LP/HC and LF/HC. Like Clegg & Shafat (2010), the results showed that the LF/LP/HC test meal reduced hunger compared to a HF/LP/LC meal (p=0.03), moreover no difference was found between the LF/HC and the HF/LP/LC test meals. Besides this a LF/LP/HC meal reduced hunger compared to a LF/HC meal (p<0.01) in experiment 2, but no difference was found in experiment 1. Cotton et al. (1994) used a basis breakfast for all test meals; this basis meal was then adjusted according to the wished macronutrient composition. The adjustment resulted in difference within the test meals quantity, where the LF/LP/HC was heavier than the other test meals. Furthermore the LF/LP/HC test meals contained more calories than the LF/HC. This difference in calories and quantity might be the reason for the reduction of hunger in the LF/LP/HC test meal. This is supported in the study by Kral (2006) who found that the portion size influence satiety. Noteworthy is that the LF/LP/HC test meal had different effects on hunger, when compared with the LF/HC test meal, in experiment 1 and 2. The two experiments had the same meals and same procedures, however the ad libitum meal was served at different time frames. In experiment 1, the meal was consumed 4 hours and 20 min after the test meal, whereas in experiment 2 it was consumed 2.5 hours after the test meal. The main difference between the compared test meals was that one was low in protein while the other had a moderate protein content. This indicates that the protein content might have an effect on how long a HC meal reduces hunger. This is in agreement with the review by Paddon-Jones et al. (2008), who found that even a moderate increase of protein might promote satiety. Similar results 34 were found for fullness and SM intake, indicating that there is a correlation between the three factors. With this Cotton et al. suggest that a LF/LP/HC meal improves satiety to a higher extent than a HF meal and a LF/HC meal, nevertheless the difference in quantity and protein cannot be disregarded as an influencing factor, as they both seem to have an influence on satiety. Dougkas & Ostman (2016) investigates a LF/LP/HC meal and compares it to HF and HP test meals. In contrast to Cotton et al. no difference was found between the LF/LP/HC and a HF/LP/LC meals. Furthermore Dougkas & Ostman (2016) found no difference between the LF/LP/HC test meal and the HF/HP/LC, LF/HP and a HP meal. Similar results were found for fullness and SM intake. The LF/LP/HC test meals used by Cotton et al. (1994) and Dougkas & Ostman (2016) has approximately the same macronutrient composition, therefore the reason for the different results must be found elsewhere. Dougkas & Ostman (2016) compared five iso-caloric and iso-volumetric liquid preloads and found no difference between the test meals. The explanation for the different result in the two studies might therefore be explained by the quantity and calorie content, which differed in the test meals by Cotton et al. (1994) but not in Dougkas & Ostman (2016). Overall the results by Dougkas & Ostman (2016) suggest that the macronutrient composition have no effect on satiety. Furthermore it indicates, as Cotton et al. (1994), that there is a correlation between hunger, fullness and SM intake. Fallaize et al. (2013) used a HC meal and similar to Dougkas & Ostman (2016), they found no significant difference between a HC meal and a HF/HP/LC. The same applied for the HF/LP test meal; similar results were found for fullness and SM intake. Fallaize et al. (2013) compared three iso-caloric test meals with different macronutrient compositions. The HC test meal was composed of cereals, the HF/HP/LC of eggs and the HF/LP of croissant. Interestingly no significant difference was found between the HC test meal and the two HF test meals, even though the satiety index score (SI%) of the foods, made by Holt et al (1995), are rated differently. HC: cereal has a SI% of 118±19, the HF/HP/LC: eggs has a SI% of 150 ±31 and HF/LP: croissant has a SI% of 47 ±17 (Holt et al., 1995). According to these satiety index scores it seems that the quantity of the test meal should be higher for the HC test meal than the HF/HP/LC test meal, in order to reach the same satiety rate. Furthermore a lower quantity should be eaten of the HC meal, in order to achieve the same satiety rating as HF/LP. Unfortunately there is no data on the quantity of the test meals and the quantity can therefore not be excluded or included as an influencing factor. The results by Fallaize et al. (2013) suggest there is no difference in satiety when comparing a HC meal 35 with HF/HP/LC and HF/LP meals. Additionally it indicates that there is a correlation between hunger, fullness and SM intake, as Cotton et al. (1994) and Dougkas & Ostman (2016). Fischer et al. (2004) investigated LF/LP/HC test meal and found that a LF/LP/HC meal reduced hunger to a lower extent than a LF//HP/LC (p<0.001). The preloads used by Fischer et al. (2004) consisted only of one macronutrient and were iso-caloric and iso-volumetric. These results indicates that a protein meal reduce hunger to an higher extent than carbohydrate, this is in agreements with the findings by Halton & Hu (2004), who found convincing evidence supporting protein increases satiety compared to the other macronutrients. Interestingly this result is in contrast with the findings by Dougkas & Ostman (2016), as it indicates that the macronutrient distribution might have an effect on satiety. The reason behind this could be that Dougkas & Ostman (2016) did not use pure macronutrients in their preloads, however the boundary for when the macronutrient distribution have an effect is unknown. Interestingly there was no difference between the HC test meal and HP test meal in the SM intake, indicating that the effect on satiety only exist within a 3hour period. The results by Fischer et al. (2004) suggest that HC meals have a lower effect on satiety than HP, furthermore it indicate that there is no correlation between hunger and SM intake, within this time frame, as seen in Clegg & Shafat (2010). Perrigue et al. (2009) used a LF/LP/HC meal when investigating carbohydrates influence on satiety. In agreement with Fischer et al. (2004), Perrigue et al. (2009) found that a LF/LP/HC meal reduced hunger less than a LF/HP/LC meal (p<0.05). Furthermore they found the LF/LP/HC meal to reduce hunger less than a LF/HFi, LF and a LF/HP/LC/HFi meal (p<0.05). Similar results were found for fullness. The main difference between the LF/LP/HC test meal and the other test meals is that the LF/LP/HC is orange juice, whereas the others are liquid yogurts. Since all the test meals is in liquid form, gastric emptying should not be the reason for the difference in hunger ratings; as gastric emptying of pure liquid occurs in the same rate and it must therefore be the nutrients in the liquids, that contribute to the difference of hunger. Another factor that could influence the difference in hunger ratings are the meal type and the pleasantness of the food, as the subjects are affected by both sensory and cognitive cues, as well as hunger (Kral, 2006). The result of Perrigue et al. (2009) therefore indicates that high protein and high fiber meals reduce hunger to a higher extent than high carbohydrate meals. Furthermore it indicates that there is a correlation between hunger and fullness. However no difference was found between the test meals and SM intake, indicating that the effect of satiety only occurs within a 2-hour range, this is in agreement with the findings by Clegg & Shafat (2010) and Fischer et al. (2004). 36 Poppitt et al. (1998) found no difference between a LF/HC meal and a LF/HP/LC or a HF/LC meal. This differentiates from the results of Fischer et al. (2004) and Perrigue et al. (2009), who found the LF/LP/HC to reduce hunger to a lower extent than LF/HP/LC test meal. This can be due to several factors, first Poppitt et al. (1998) uses solid food, whereas the others uses liquids, resulting in a longer gastric emptying for the solid foods, as the lag phase extent this process. Secondly the test meal by Poppitt et al. (1998) has a moderate protein content, which has shown a tendency for higher satiety (Paddon-Jones et al., 2008). Similar results were found in fullness ratings, indicating that there is a correlation between hunger and fullness. Comparable results were found for the SM intake, however one result contradicts the others. The LF/HC test meal resulted in a higher intake compared to the LF/HP/LC test meal. This can be due to sensory specific aspects, as the HC test meals evoked the appetite to a higher extent than the LF/HP/LC, resulting in a higher intake at the SM. Related findings are found in the study by Kral (2006), when investigating sensory specific effect on SM intake. Overall the results by Poppitt et al. (1998) suggest that there is a correlation between hunger and fullness, but not to SM intake. Similar to Poppitt et al. (1998), Rolls et al. (1991) found no difference between a HF/LC meal and the LF/HC test meal. The protein content for the two test meals were identical, therefore only the carbohydrate and fat content differed, showing that there is no variation between high fat and high carbohydrate meals. Alike results were found for SM intake indicating that there is a correlation between hunger, fullness and SM intake, as the findings by Cotton et al. (1994), Dougkas & Ostman (2016) and Fallaize et al. (2013). Rolls et al. (1998) investigated two HC meals: LF/LP/HC and LP/HC. In agreement with Dougkas & Ostman (2016), no difference was found between a LF/LP/HC meal and a HP meal. Furthermore they found that a LF/LP/HC meal reduced hunger significantly more than a HF/LP meal, a LP and a LP/HC meal (p<0.01). Similar results were found for fullness and SM intake. Rolls et al. (1998) used iso-caloric test meals. The LF/LP/HC test meal was composed of pasta and the HP of chicken, no difference was found between these test meals, which is in agreement with the SI% by Holt et al. (1995), where carbohydrate rich foods SI% are 166±24 and protein rich foods SI% are 166±13. The LF/LP/HC test meal effect on hunger was rated higher than the rest of the test meals, which can be due to several factors. First the other test meals were snack foods, whereas the LF/LP/HC meal was a main dish. Secondly snack foods have in general a low SI% at 100±10 (Ibid). Lastly the quantity of the other test meals differed, as the LF/LP/HC test meal weighed more than the HF/LP, LP/HC and LP test meal. The results for the LP/HC test meal 37 showed that a LP/HC reduced hunger to a lower extent than a HP, HF/LP and a LP meal. The findings are in contrast with results for the LF/LP/HC test meal. One reason for this difference can be the quantity of the LF/HC that is much lower; this is supported by Kral (2006), who found that the quantity matter. Another reason could be the difference in fat content in the two HC test meals, indicating that a moderate fat content in a HC meal might result in a lesser reduction of hunger compared to a HC meal low in fat. The results for the LP/HC test meal in fullness and SM intake differed for those in hunger, indicating there is no correlation between hunger, fullness and SM intake, within the time period. In contrast to the hunger ratings no difference was found on fullness between the HF/LP and LP test meal compared to the LP/HC test meal. This can be due to different quantity of the foods, thus the rates of gastric emptying of the specific food become somewhat similar. The results from Rolls et al. (1998) indicate that the correlation between hunger, fullness and SM intake might be influenced by the macronutrient composition and quantity. Carbohydrates influence on satiety is inconsistent. The majority of the studies found no difference on the carbohydrate test meals, compared to the others. However there seems to be inconsistency in the carbohydrates effect, within the different types of carbohydrate meals. This indicates that factors, as macronutrient composition, meal type, quantity and calorie content, might effect the satiation of the test meal. The studies that found a significant difference between the test meals showed that a high carbohydrate meal is less satiating than protein, as seen in the studies by Fischer et al. (2004), Perrigue et al. (2009) and Rolls et al. (1998). Moreover the results showed a tendency toward fat being less satiating than carbohydrates, as found by Rolls et al. (1998), Cotton et al. (1994) and Clegg & Shafat (2010). Lastly Perrigue et al. (2009) found carbohydrates to be less satiating than HFi. Fiber & satiety Two studies investigated high fiber meals effect on satiety; therefore this thesis has insufficient studies, in order to conclude on the effect of fiber and satiety. However the studies included can somewhat show a tendency of high fiber meals effect on satiety. Kristensen et al. (2010) compared a HFi meal with HP meals. No differences were found between the HFi test meal and the HP test meals within hunger. Similar results were found for fullness and the SM intake. Comparable results were found, in the study by Holt, Brand-Miller, & Stitt (2001), when comparing breads with different fiber contents. The results by Kristensen et al. 38 (2010) indicates that HFi and HP have the same effect on satiety, moreover it suggests that there is a correlation between hunger, fullness and SM intake. Perrigue et al. (2009) tested two HFi meals effect on satiety: LF/HFi and LF/HP/LC/HFi. Likewise Kristensen et al. (2010), Perrigue et al. (2009) found no difference between the ratings of hunger, after the HFi test meals and the LF/HP/LC test meal. Moreover they found no difference between the hunger ratings after the HFi test meals and the LF or between the HFi test meals. Lastly the LF/LP/HC test meal resulted in higher hunger ratings than the two HFi test meals (p<0.05). This difference in hunger ratings may be due to the digestion process of fibers, as it does not begin before it reaches the large intestine, thereby extending the process. The results for fullness were somewhat similar to those of hunger, however in contrast to Kristensen et al. (2010) the fiber test meals were found to result in a higher rating of fullness compared to the LF/HP/LC test meal (p<0.05). Several factors could be the reason for the discrepancy. First Perrigue et al. (2009) serves liquid foods whereas, Kristensen et al. (2010) serves solid foods, resulting in a faster gastric emptying process, as the liquids are not effected by the lag phase. Hereby only the liquids are emptied at approximately the same rate, as the nutrient content is similar; thereby only the liquids containing fiber are affected by the body’s inability to digest fibers. Secondly the fat content might have an influence as Perrigue et al. (2009) test meals contained a lower fat content compared to Kristensen et al. (2010). Thirdly the calorie content might have a impact on fullness, as Perrigue et al. (2009) HP test meal contained fewer kcal than the HFi test meals, whereas the calorie content were constant in Kristensen et al. (2010) test meals. Lastly no difference was found between the different test meals and the SM intake, indicating that the satiating effect only occur within a 2-hour range. Overall the results by Perrigue et al. (2009) suggest that there is no difference on satiety between HP and HFi meals, moreover the HFi results in higher satiety compared to HC test meals. Macronutrients & satiety The literature showed a tendency for protein to be more satiating than fat and carbohydrate, where carbohydrate was superior to fat. In addition it would seem that fiber was more effective than carbohydrate in inducing satiety. The reasons for this effect are complex and remain to be explored; several factors may be of influence, both physiological and psychological. Even though it would seem that calorie content of a meal is of weak influence in regards to satiety, energy density may still be a factor that could be of influence to the outcome (Poppitt et al., 1998). The digestion, rates 39 of pre-absorptive and post-absorptive mechanisms of fat, protein, carbohydrate and fiber may also be a considerable factor, as it differs considerably and could cause the different short-term satiety effects (Fischer et al., 2004). Furthermore subjective opinions on palatability caused by sensory properties such as: feel, viscosity, touch, taste and moisture may explain the conflicting results within the studies, as palatability previously has been found to affect energy intake. This is seen in the studies by Dougkas & Ostman (2016), Rolls et al. (1991), Rolls et al. (1988) and Rolls et al. (1988). Such a hypothesis is supported by Buckland et al. (2015), who found subjective opines on satiety of different foods. Lastly diet induced thermogenesis has been associated with satiety (Westerterp-Plantenga et al., 1997). There is indicative evidence towards diet induced thermogenesis having an effect on satiety scores; the higher the diet-induced thermogenesis is, the more satiating a meal is. In this context full fat meals have the lowest diet induced thermogenesis compared to protein as the highest (Crovetti, Porrini, Santangelo, & Testolin, 1998). Protein has in previous research proved to be more satiating than the other macronutrients, thereby limiting the energy intake after a meal, while also increasing thermogenesis; influencing satiety and energy expenditure (Paddon-Jones et al., 2008). An important factor to consider is the different procedures implemented in the studies, more specifically at what time the different parameters such as hunger were measured. Time seems to be of the essence, concerning satiety and its related factors. The studies by Fallaize et al. (2013), Cotton et al. (1994) and Clegg & Shafat (2010), who measured on whole day satiety found no difference on satiety parameters, suggesting that satiety is limited to a short time frame; <7-10 hours. In addition differences between satiety measures were limited to ratings prior to the ad libitum meal, similarly suggesting that the SM overwrites the effect of a previous meal. This proposes that to obtain the full effect of macronutrient induced satiety, one would be recommended to follow the same effective macronutrient composition throughout the day. Cotton et al. (1994) and Fischer et al. (2004) found a significant difference on hunger ratings following the test meals, yet no effect in the SM intake. A reason for the given outcome may be due to the extended time between a test meal and the SM, suggesting that satiety only affects subsequent intake within a ≤3-4 hour period. Perrigue et al. (2009) was the only study, which found a difference on hunger, but not on SM intake within a <3-hour period; this was likely due to the meal type presented in the study. Studies including a protocol concerning a 3-hour period, found a significant difference on hunger ratings and SM intake; supporting the findings from Cotton et al. (1994) and Fischer et al. (2004), that the effect of satiety occurs within a 3-hour timeframe. 40 5.1 Literature considerations The limitations and strength of the included studies needs to be taking into consideration in order to evaluate the results of this thesis. It is important to recognize that it can be difficult to conduct research on the control of energy intake, as it is both expensive and complicated to measure habitual intake (Geissler & Powers, 2005, p. 96). Something to remember is therefore that all the included studies tested short-term satiety parameters, which means that the outcome they found may not be of relevance for long-term interventions. This would have to be investigated in studies to come. All the included studies are randomized control trials (RCT), which has the strength of avoiding selection bias (Carneiro & Howard, 2011, p. 122), moreover RCT is rated among the highest in the evidence hierarchy (Kjøller et al., 2007, p. 371). Two different trial designs are used in the included studies; crossover and factorial trials. Furthermore all the included studies implemented crossover designs, where the subjects' acts as it own control (Thelle, 2015, p. 131). Benefits of using own control, in studies dealing with satiety, are that the rating of satiety is subjective. Therefore by using the control of the subjects rating of hunger, at a fasting phase, will reduce possible bias, as it reflects the subject’s satiety ratings on the specific test day. Crossover trials are appropriate for short-term studies, as the subjects have a washout period between each test meal. This enables to control for confounders and make alterations (Carneiro & Howard, 2011, p. 124). Cotton et al. (1994), Dougkas & Ostman (2016) and Perrigue et al. (2009) also included a factorial design, where the results are both compared with the subjects own control and a control meal (Thelle, 2015, p. 131). This design has the benefit of assessing interactions between the test meals and the subjects own control, thereby evaluating on various interventions at once, which saves time and money (Carneiro & Howard, 2011, p. 123). Lastly all the studies, except the one by Kristensen et al. (2010), are single-blinded. This has the strength of limiting information bias from the subjects in the study (Carneiro & Howard, 2011, p. 127). The study design could have been even stronger if double-blinding was used, hereby the information bias from the observer also could have been reduced (Ibid). Kristensen et al (2010) uses an open-labeled study, which should be taking into consideration when evaluating the results. In relation to study design, an important component is the number of included participants. This is an important factor, which can determine the outcome of a study; studies with insufficient power may fail to find any effect on the different treatments (Flint et al., 2000). Our thesis included studies with as few as 9 participants (Clegg & Shafat, 2010) to as many as 40 (Perrigue et al., 2009); the statistical power of the different studies is therefore of relative variation. 41 However this limitation can be difficult to avoid, due to the extent of work required for these types of studies. In addition some of the included literature did experience dropouts due to: poor compliance, being unable to attend study days, personal reasons, protocol non-adherence, logistic problems and dislike of the ad libitum meal. Dropouts were limited to 5 studies; varying from 3% dropout (Fallaize et al., 2013) to 20% (Kristensen et al., 2010). In regards to participants, few studies took gender differences into account. This is an issue that must be considered, as there may be differences between the sexes on satiety regulation, for example in relation to portion size (Perrigue et al., 2009). As men are of larger body size and energy needs, they would require larger portions to experience the same rate of satiety. If not taken into account, within the different studies, this factor could result in large variation, thereby hiding the real effect (Dougkas & Ostman, 2016). Evaluating the rating scales and thereby the measurements, several factors must be considered. First of all, one should keep in mind that the measurements are subjective; therefore the subjects might interpret the feeling of hunger and fullness differently. However the subjects act as their own control, thereby minimizing possible bias from the different interpretations. Secondly, different study parameter must be considered, as the included subjects might alter the results or require different study designs, such as prior diet standardization (Flint et al., 2000). The included studies by Dougkas & Ostman (2016), Fischer et al. (2004), Perrigue et al. (2009), and Poppitt et al. (1998) uses prior diet standardization, in order to reduce possible bias on the rating of satiety. However the fact that the other studies do not include prior diet standardization should not be viewed as a limitation, due to the fact that it does not seem to affect VAS ratings (Flint et al., 2000). Other factors that have been shown to affect the ratings are age, gender, and physical activity (Gregersen et al., 2011). These influencing factors are all taken into consideration, in order to minimize possible bias. All the studies subjects are within the same age range of 20-30 years, thereby not affected by the age influence on satiety ratings (Ibid), making the included studies comparable. The different satiety ratings between genders, are likely due to hormonal differences (Ibid), thus it is difficult to take into account, when planning the procedure of the study. Dougkas & Ostman (2016) and Rolls et al. (1991) made alterations in the test meals calorie content and quantity, in order to make up for the gender differences. Furthermore all the included studies instructed the subjects to refrain from strenuous physical activity the day before the test, to avoid alterations in the results. A third factor affecting the results is the subject’s perception of a meal’s satiety. Previous studies have found that the foods energy density, fiber content, macronutrient 42 composition and perceived healthiness are associated with the perceived satiety (Buckland et al., 2015). As the subjects included in the studies are able to see the food they consume, it cannot be overruled that the ratings of satiety are influenced by the perceived satiety, combined with the satiety effect from the food consumed. There are several limitations considering the protocols of the included studies, which in some way or another not only limit the comparability between the different studies, but also between the different test meals. Within this consideration there are three primary factors, which differ between test meals and seems to have an influence on satiety; quantity/portion size, test meal structure and meal type. Portion size seems to be a primary factor influencing satiety (Kral, 2006). Some studies compare test meals with up to a 6 times difference in quantity (Rolls et al., 1988), this makes it difficult to determine whether the macronutrient composition or the quantity is the cause for the given effect. Other test meals are presented in liquid form, while most are served as solid foods. There is however suggestive evidence for the test meal structure having an influence on satiety ratings (Dougkas & Ostman, 2016), which can limit comparison between different test meals. Last but not least, the test meals did come in different meal types, as in the study by Rolls et al. (1988), where for example chicken and pasta were served, which many would consider as a dinner and turkish delight and confectionery, which is often presented as snacks or dessert. The attribute of not presenting typical breakfast foods, may influence the subjective perception of a test meals satiety value and thereby affecting the results (Dougkas & Ostman, 2016). These limitations must be taken into consideration when viewing the results of the included studies, as there are several confounding factors, which hinder clear conclusions. There seems to be a lack of mentioning clinical significance. Statistical significance is of high importance, in order to see if the findings are caused by chance or due to treatment (Valanis, 1999, p. 363). Statistical significance does however not necessarily speak for the clinical significance: whether or not the outcome is of real importance in treatment. The statistically significant effect may not be big enough for the treatment to have an actual clinical effect, which could influence a given outcome (Ibid). For example a high protein test meal compared to a high fat test meal might score a p-value of 0.05 on SM intake, which would be deemed as statistically significant. The clinical difference might however be less significant, if there was<10 calories between the test meals. This would most likely not have an impact on weight control, and would therefore not be considered clinically important. 43 6. Conclusion The association between macronutrients and their effect of satiety was inconclusive, as further influencing factors were established. Quantity/portion size, meal type and sensory cues all seemed to play an important role in the rating of satiety. The influence of these aspects needs more research in order to examine their contribution to satiety. However this thesis found suggestive evidence for fat to affect satiety to a lesser extent than both carbohydrate and protein, thereby it seems that high fat meals increase energy intake. In addition protein seems to increase fullness while decreasing hunger and subsequent meal intake, due to its increased satiating effect. Carbohydrates effect on satiety is inconsistent, however a weak tendency points toward carbohydrates being more satisfying than fat and less satisfying compared to protein. Lastly the effect of fiber seems to result in a higher satiety compared to carbohydrates. The reason for the given outcome is still not known and may be caused by both physiological and psychological factors. 7. Perspectives This thesis investigated how macronutrients invoke satiety and to what degree. We found suggestive evidence for how to compose the macronutrient composition of a meal, in order to increase satiety and limit subsequent intake. This information may provide an effective tool for future counseling and become of great value in the fight against obesity. In the examination, this thesis came across several factors that may have an influence on satiety. Therefore more research is required, in order to understand satiety and make it a useable tool in fighting overweight and obesity. The subjective perception of different foods and their satiating abilities was found to be an interesting factor that may influence satiety. Buckland et al. (2015) supports this hypothesis, as their results of perceived satiety of foods, showed the perceptions of food to be of influence. To what extend and why perceived satiety influences satiety ratings, is to our knowledge unknown. More research is therefore needed in this area, as it may be of significant importance in relation to satiety. Earlier studies have previously investigated satiety in regards to a specific value (SI%). Holt et al. (1995) found foods to be an important factor, when investigating satiety, as foods within different macronutrient groups, differ in their satiety score. It would therefore be of great value to investigate, which foods have the greatest satiety rate and combine it within the macronutrient composition of a meal. It is important to chain the foods with the macronutrients composition, as 44 single foods cannot ensure overall health; the whole diet need to taken into consideration (Nordic Council of Ministers, 2008, p. 21). The quantity was also found to be an influencing factor in this thesis; therefore we recommend future studies to investigate the portions size effect on energy intake, as this angle could be essential in relation to satiety, when fighting weight problems. In general there is still many relations and questions that may be worth investigating, when looking at satiety. The successful treatment of obesity and weight gain may very well depend on research to come. 8. Practice considerations Every year, a substantial amount of people deal with overweight and weight gain, this tendency causes millions of deaths (WHO, 2014, p. 79). There are clear indications for the primary causes, providing this tendency, being increased food intake, lack of physical activity and an abundance of food availability (Aljuraiban et al., 2015). However when we know the reason for the given problem, how are we not able to fix it? Weight gain and weight control is a much more complex and thorny subject, than what most people believe it to be. The problem is not only influenced by physiological elements, but also psychological; this combination makes it extremely difficult to lose weight after it has been gained (Byrd-Bredbenner et al., 2012, pp. 342–342). In line with the increasing knowledge availability in the society, the general population finds it more difficult to distinguish between right and wrong. This has resulted in people trying different diets, often with no support of scientific evidence. Instead people choose diets recommended by friends, families or celebrities, who has had a positive experience with the diet; FAD diets has become relatively popular, yet they rarely provide any long-term success (Byrd-Bredbenner et al., 2012, pp. 338–341). But instead of focusing on how to resolve the problem, it might be beneficial to know how to prevent it in the first place. It would seem that one of the primary causes, for causing weight gain, is increased food intake. A relevant question would therefore be; how can we effectively limit this tendency? A perspective could be through satiety: by increasing the satiety of our habitual diets, we may be influenced to eat less. It is important for our profession to continue research in the field of health, thus we may be able to find evidence based and effective solutions to the problems of our modern society. Through our thesis we investigated satiety and more specifically how it might be influenced by the different 45 macronutrients. We found suggestive evidence for increasing satiety within a meal, by increasing protein and fiber content while decreasing fat. With the given finds, we may be able to implement a whole day diet consisting of the stated macronutrient composition that could increase satiety and limit subsequent intake throughout the day. With this tool, we may provide clients, as well as the population, an effective way of controlling their weight and thereby avoiding weight gain in the first place. In addition such a diet tool may increase adherence for people trying to loose weight, through a restricted diet. This new information may help us, as health professionals, to guide our clients to a better and healthier life. 46 9. List of reference Aljuraiban, G. S., Chan, Q., Oude Griep, L. M., Brown, I. J., Daviglus, M. L., Stamler, J., … Frost, G. S. (2015). The Impact of Eating Frequency and Time of Intake on Nutrient Quality and Body Mass Index: The INTERMAP Study, a Population-Based Study. Journal of the Academy of Nutrition and Dietetics, 115(4), 528–536. Aller, E. E. J. G., Larsen, T. M., Claus, H., Lindroos, A. K., Kafatos, A., Pfeiffer, A., … van Baak, M. A. (2014). Weight loss maintenance in overweight subjects on ad libitum diets with high or low protein content and glycemic index: the DIOGENES trial 12-month results. International Journal of Obesity, 38(12), 1511–1517. Astrup, A., Bügel, S., dyerberg, J., & Sten, S. (Eds.). (2015). Menneskets ernæring (4th ed.). Kbh.: Munksgaard. Blundell, J. E., & Burley, V. J. (1987). Satiation, satiety and the action of fibre on food intake. International Journal of Obesity, 11(1), 1–92. Buckland, N. J., James Stubbs, R., & Finlayson, G. (2015). Towards a satiety map of common foods: Associations between perceived satiety value of 100 foods and their objective and subjective attributes. Physiology & Behavior, 152, 340–346. Byrd-Bredbenner, C., Moe, G., Beshgetoor, D., & Berning, J. (2012). Wardlaw’s perspectives in nutrition. (9th ed.). New York; London: McGraw-Hill Higher Education. Carneiro, I., & Howard, N. (2011). Introduction to Epidemiology (Second edition). Maidenhead: Open University Press. Clegg, M., & Shafat, A. (2010). Energy and macronutrient composition of breakfast affect gastric emptying of lunch and subsequent food intake, satiety and satiation. Appetite, 54(3), 517–523. Cotton, J. R., Burley, V. J., Weststrate, J. A., & Blundell, J. E. (1994). Dietary fat and appetite: similarities and differences in the satiating effect of meals supplemented with either fat or carbohydrate. Journal of Human Nutrition and Dietetics, 7, 11–24. Crovetti, R., Porrini, M., Santangelo, A., & Testolin, G. (1998). The influence of thermic effect of food on satiety. European Journal of Clinical Nutrition, 52, 482–488. Dougkas, A., & Ostman, E. (2016). Protein-Enriched Liquid Preloads Varying in Macronutrient Content Modulate Appetite and Appetite-Regulating Hormones in Healthy Adults. Journal of Nutrition, 146(3), 637–645. Evans, D. (2003). Hierarchy of evidence: a framework for ranking evidence evaluating healthcare interventions. Journal of Clinical Nursing, 12(1), 77–84. Fallaize, R., Wilson, L., Gray, J., Morgan, L. M., & Griffin, B. A. (2013). Variation in the effects of three different breakfast meals on subjective satiety and subsequent intake of energy at lunch and evening meal. European Journal of Nutrition, 52, 1353–1359. Fischer, K., Colombani, P. C., & Wenk, C. (2004). Metabolic and cognitive coefficients in the development of hunger sensations after pure macronutrient ingestion in the morning. Appetite, 42(1), 49–61. Flint, A., Raben, A., Blundell, J. E., & Astrup, A. (2000). Reproducibility, power and validity of visual analogue scales in assesment of appetite sensations in single test meal studies. International Journal of Obesity, 24, 38–48. Fogelholm, M., Anderssen, S., Gunnarsdottir, I., & Lahti-Koski, M. (2012). Dietary macronutrients and food consumption as determinants of long-term weight change in adult populations: a systematic literature review. Food & Nutrition Research, 56(0). Geissler, C., & Powers, H. J. (Eds.). (2005). Human nutrition (11th ed). Edinburgh ; New York: Elsevier/Churchill Livingstone. Gerstein, D. E., Woodward-Lopez, G., Evans, A. E., Kelsey, K., & Drewnowski, A. (2004). Clarifying concepts about macronutrients’ effects on satiation and satiety. Journal of the American Dietetic Association, 104(7), 1151–1153. Gregersen, N. T., Møller, B. K., Raben, A., Kristensen, S. T., Holm, L., Flint, A., & Astrup, A. (2011). Determinants of appetite ratings: the role of age, gender, BMI, physical activity, smoking habits, and diet/weight concern. Food & Nutrition Research, 55(00). Halton, T. L., & Hu, F. B. (2004). The Effects of High Protein Diets on Thermogenesis, Satiety and Weight Loss: A Critical Review. Journal of the American College of Nutrition, 23(5), 373–385. Hjørland, B. (2011). Evidence-based practice: An analysis based on the philosophy of science. Journal of the American Society for Information Science and Technology, 62(7), 1301–1310. Holm, A. B. (2013). Philosophy of science: an introduction for future knowledge workers. Frederiksberg C: Samfundslitteratur. Holt, S., Brand Miller, J., Petocz, P., & Farmakalidis, E. (1995). A satiety index of common foods. European Journal of Clinical Nutrition, 49, 675–690. Holt, S. H. A., Brand-Miller, J. C., & Stitt, P. A. (2001). The effects pf equal-energy protions of different breads on blood glucose levels, feelings of fullness and subesequent food intake. Journal of the American Dietetic Association, 101, 767–773. Kjøller, M., Juel, K., & Kamper-Jørgen, F. (Eds.). (2007). Folkesundhedsrapporten Danmark 2007 (2. opl). København: Statens Institut for Folkesundhed. Kral, T. V. E. (2006). Effects on hunger and satiety, perceived portion size and pleasantness of taste of varying the portion size of foods: A brief review of selected studies. Appetite, 46(1), 103–105. Kristensen, M., Jensen, M. G., Riboldi, G., Petronio, M., Bügel, S., Toubro, S., … Astrup, A. (2010). Wholegrain vs. refined wheat bread and pasta. Effect on postprandial glycemia, appetite, and subsequent ad libitum energy intake in young healthy adults. Appetite, 54(1), 163–169. Larsen, T. M., Dalskov, S.-M., van Baak, M., Jebb, S. A., Papadaki, A., Pfeiffer, A. F., … Astrup, A. (2010). Diets with high or low protein content and glycemic index for weight-loss maintenance. New England Journal of Medicine, 363(22), 2102–2113. Layman, D. K., Evans, E. M., Erickson, D., Weber, J., Bagshaw, D., Griel, A., … Kris-Etherton, P. (2009). A moderate-protein diet produces sustained weight loss and long-term changes in body composition and blood lipids in obese adults. The Journal of Nutrition, 139(3), 514–521. Longenbaker, S. N. (2014). Mader’s understanding human anatomy & physiology (8th ed.). Murphy, K. G., & Bloom, S. R. (2006). Gut hormones and the regulation of energy homeostasis. Nature, 444(7121), 854–859. NCBI. (n.d.-a). Adult - MeSH - NCBI. Retrieved November 8, 2016, from https://www.ncbi.nlm.nih.gov/mesh/68000328 NCBI. (n.d.-b). Satiation - MeSH - NCBI. Retrieved November 8, 2016, from https://www.ncbi.nlm.nih.gov/mesh/?term=satiation Nordic Council of Ministers, N. C. of M. (2008). Nordic Nutrition Recommendations 2012. Nordic Nutrition Recommendations 2012, 5(11), 1–3. Paddon-Jones, D., Westman, E., Mattes, R. D., Wolfe, R. R., Astrup, A., & Westerterp-Plantenga, M. (2008). Protein, weight mangement, and satiety. American Journal of Clinical Nutrition, 87, 1558S–61S. Perrigue, M. M., Monsivais, P., & Drewnowski, A. (2009). Added Soluble Fiber Enhances the Satiating Power of Low-Energy-Density Liquid Yogurts. Journal of the American Dietetic Association, 109(11), 1862–1868. Poppitt, S. D., Mccormack, D., & Buffenstein, R. (1998). Short-term effects of Macronutrient Preloads on Appetite and Energy Intake in Lean Women. Physiology & Behavior, 64(3), 279–285. Rolls, B. J., Hetherington, M., & Burley, V. J. (1988). The specificity of satiety: the influence of foods of different macronutrient content on the development of satiety. Physiology & Behavior, 43, 145–153. Rolls, B. J., Kim, S., Mcnelis, A. L., Fischman, M. W., Foltin, R. W., & Moran, T. H. (1991). Time course of effects of preloads high in fat or carbohydrate on food intake and hunger ratings in humans. American Journal of Physiology, 4, 756–762. Shils, M. E., Shike, M., Ross, A. C., Caballero, B., & Cousins, R. (Eds.). (2006). Modern nutrition in health and disease (10th ed). Philadelphia: Lippincott Williams & Wilkins. Thelle, D. (2015). Epidemiology: A basis for public health and disease prevention. Oslo: Gyldendal. Valanis, B. (1999). Epidemiology in health care (3rd ed). Stamford, Conn: Appleton & Lange. Westerterp-Plantenga, M. S., Wijckmans-Duijsens, N. E., Verboeket-Van De Venne, W. P., De Graaf, K., Weststrate, J. A., & Hof, K. H. V. H. (1997). Diet-induced thermogenesis and satiety in humans after full-fat and reduced-fat meals. Physiology & Behavior, 61(2), 343–349. WHO. (2014). Global status report on noncommunicable diseases 2014: attaining the nine global noncommunicable diseases targets; a shared responsibility. Geneva: World Health Organization.