The oat oil fraction used in these experiments was high in polar lipids and obtained by controlling the concentration of ethanol, water, and sugar in different fractionation steps after ethanol extraction of oats. The lipid class composition of the fractionated oat oil was determined by HPLC (22). The main lipid classes are presented in Table 1.

Very small, stable, and uniform liposomes were formed almost spontaneously during dilution of the fractionated oat oil in a carefully controlled composition of ethanol–water mixture. The liposomes formed were diluted further in water and finally the ethanol was evaporated (19). The final liposome preparation contained 10 wt% lipids and 90 wt% water. The volumetric average diameter of the liposomes was 100 nm, measured by dynamic light scattering (Zetasizer Nanoseries, Malvern Instruments).

The stability of the liposomes was investigated by dilution 1:1 in: water; citric acid buffers pH 4, pH 3; citric acid 1 M. It was not possible to detect any changes in structure after 4 months in water. The stability decreased at decreasing pH. Using 1 M citric acid, it lasted 4 days before any changes could be seen. Thus, it is very reasonable that these liposomes were sufficiently stable to survive the low pH in the stomach and that they could enter the duodenum intact.

In the first part, 19 healthy men and women (10 women, 9 men) of mean age 42, BMI 25, were recruited through announcements on the local intranet and advertisements. On two occasions 10–14 days apart, they were randomly assigned either a test- or placebo breakfast. In the second part of the study, we enrolled 15 women of mean age 34 and BMI 24. On three occasions, the 15 women received two doses of LOO in test breakfast or yoghurt in the placebo breakfast.

Exclusion criteria were abnormal blood glucose levels in the fasting state, and hyperlipidaemia in addition to signs of liver or kidney disease or inflammatory activity as indicated by raised C-reactive protein, liver tests (bilirubin, alkaline phosphatase, gamma-glutamyl transpeptidase, and aminotransferases), and creatinine levels. All participants gave their written informed consent prior to the study. The studies were approved by the Regional Human Ethics Committee of Lund-Malmö, Sweden (Dnr 2010/18, 2011/55).

Participants conducted food registrations in the form of diaries over 4 days before the experiments and also of the remainder of the day of each trial. The intake was analyzed by the use of Dietist XP software with the Swedish National Food Administration database (2008 03 06) (Kost och Näringsdata, Sweden). Dietary intakes were related to recommended Swedish dietary guidelines (24).

The breakfast meals were composed of one slice of bread (37 g), one slice of smoked ham with 3% fat (14 g), blackberry jam (40 g), coffee or tea without milk or sugar and free access to water. The blackberry jam was mixed into the placebo and test meal before serving in order to mask any difference in taste and appearance. Both test- and placebo yoghurt contained 35 g of lipids at a concentration of 10 wt%. The placebo yoghurt was plain non-flavored dairy yoghurt with 10% fat. The test yoghurt contained three different amounts of LOO containing 35 g lipids in the first study, and 14 and 1.8 g lipids in the follow-up study. The remaining lipids up to 35 g were from milk; the total fat concentration was 10%. The fatty acid composition of the two lipid sources is presented in Table 2 and the lipid classes are presented in Table 1. A fat-free milk powder was added to the LOO in order to balance intake of protein, calcium, and carbohydrate in the meal (Table 3).

Participants arrived at 7.00 am. Blood samples from the fore arm vein were taken before breakfast, when the subjects had been fasting since 9.00 pm the night before. The succeeding blood samples were taken 1, 3, 5, and 7 h after breakfast. All blood samples were immediately centrifuged and plasma was kept at −70°C until analysis. The participants were not allowed to eat anything for the following 5 h after breakfast. After blood sampling at 5 h after breakfast, the participants were served ad libitum a selection of pre-weighed food items (bread, egg, tomato, red bell pepper, low-fat spread cheese, smoked ham, and a banana) and the amounts were registered. The final blood sample was taken 7 h after breakfast. Finally, the participants registered what they ate for the rest of that day.

The blood samples were analyzed for triglycerides, cholesterol, blood glucose, insulin, HDL and LDL cholesterol, liver tests, C-reactive protein, hemoglobin, white blood cell count, and platelet count. All analyses were done at the Department of Clinical Chemistry, University Hospital of Lund, Sweden using the accredited routine methods developed by Roche Diagnostics (MSDS – COBAS Integra^® – General Chemistry). Two milliliter EDTA-plasma with the addition of a protease inhibitor, aprotinin (Trasylol^®), were immediately frozen at −70°C for later analyses of non-esterified fatty acids (NEFAs) (NEFA-c kit, Wako Chemicals, USA) the gut hormones PYY, GLP-1, GLP-2, and CCK. Total PYY (3–36 and 1–36), total GLP-1 (7–36 and 9–36) and GLP-2 were analyzed by ELISA from Merck Millipore, Solna Sweden. CCK was measured by a RIA, which is entirely specific for the bioactive CCKs in plasma (26, 27).

Participants were asked to subjectively rate their feeling of fullness (satiety), hunger, and desire to eat on a 10-cm visual analog scale (VAS) at each time point of blood sampling. For example, the level of satiety was expressed as a percentage on a 10-cm scale where 100% means totally satiety and 0 means no sensation of hunger at all.

They were also asked to report any gastrointestinal sensation/side effect and notable difference between each meal regarding satiety and hunger.

Plasma values were normalized according to an averaged blank between placebo and test meal(s). Significance of differences, between the results obtained with the LOO and placebo yoghurt, was examined using Student's paired t-test or paired ANOVA multiple measures. Bonferroni post hoc test was carried out when significant differences between diets were found. The calculation of incremental area under curve (iAUC) was done using the trapezoid rule on the period 0–5 h. The Student's paired t-test was used to calculate if there were any difference between the LOO and placebo after specific time points. Linear regression analysis was performed on the correlation between caloric intake and plasma PYY concentrations. All statistical analyses and calculations of iAUC were performed with GraphPad Prism (version 5.00. GraphPad software, San Diego, CA, USA).

In the first study, 19 participants ingested 410 ml LOO containing 35 g lipids or placebo meal with an equal amount of dairy lipids (Table and 3 ). There were both men and women in both groups.

The iAUC_0–5 h for plasma TG was not significantly different after the LOO compared to placebo; however, after 7 h the TG response was significantly higher with the LOO meal (Fig. 1) (n=19, p=0.016).

The iAUC_{0–5 h} for plasma glucose was significantly decreased (n=19, p=0.03) after intake of LOO compared to placebo (Fig. 2), with a significant lower concentration of plasma glucose at all postprandial time points (n=19, p=0.01, 0.0004, 0.01, and 0.0027).

The concentration of NEFAs in plasma was slightly lower after LOO compared to after placebo at 1, 3, and 5 h (trends p<0.1). The plasma concentrations of total-, HDL-, and LDL cholesterol did not change postprandial. The minor fluctuations of the plasma concentrations of cholesterol seen after placebo were not significantly different from concentrations after LOO. We calculated the concentration of large cholesterol-containing particles (LC particles) by subtracting HDL and LDL cholesterol from total cholesterol. We found that the concentration of LC particles was more evenly distributed over time after LOO and did not display any peak concentration at 3 h as after placebo yoghurt (Fig. 3). This difference in pattern reflects a prolonged output of TG-rich lipoproteins and their remnants after a breakfast with LOO compared to the placebo yoghurt.

The concentrations of the hormones PYY, GLP-1, and CCK after intake of LOO with 35 g lipids were significantly increased compared to placebo, especially at the later time points 3, 5, and 7 h (Fig. 4a–c). The iAUC_{0–5 h} for GLP-1 and CCK was significantly increased after LOO compared to placebo (p=0.028 and 0.048 respectively). CCK was significantly increased after 3 h with LOO (p=0.019) and GLP-1 was significantly increased after 5 h (p=0.021). PYY, GLP-1, and CCK were significantly increased with LOO after 7 h compared to placebo (p=0.004, 0.014, and 0.012, respectively).

The levels of PYY, GLP-1, and CCK were significantly higher after 5 h with LOO (Figs. 4 and 7). It is reasonable that high levels of satiety hormones before lunch will lead to reduced energy intake during the rest of the day. Therefore, we investigated whether there was a correlation between the concentration of the satiety hormones just before lunch and the total energy intake on the experimental day. In Fig. 5, the results for PYY just before lunch (PYY_5 h) are plotted against the total energy intake during the experimental day for all the participants with completed data (both 35 g LOO and placebo). By using these data, we derived an equation (Eq. 1). There was a negative trend between PYY_{5 h} and total energy during the rest of the day (Spearman r -0,247, CI-0.5255 to 0.08024, p= 0.0986). When we use Eq. 1 on the PYY results from the second study in this paper (14 g LOO vs. placebo), it indicates that an intake of LOO containing 14 g lipids at breakfast will reduce the total energy intake for the whole day with 7% compared to intake of the same amount of lipids from yoghurt at breakfast. GLP-1 showed a similar trend as PYY (data not shown). 1 Energy Intak e rst of day = 2036 - 17 .72*PY Y 5th

On the VAS scale formula, women reported significantly higher satiety 7 h after intake of LOO compared to placebo, 70% satiety after LOO vs. 58% after placebo (n=10, p=0.0056). However, for men (n=9) we did not find any significant differences between the meals (data not shown).

The follow-up study on 15 women investigated the effects from intake of two lower concentrations of LOO containing 14 and 1.8 g lipids. The meals were compensated with fat-free milk powder to balance the intake of protein, calcium and carbohydrate and plain non-flavored yoghurt to balance the total amount of fat to 35 g. The composition of the meals and study design were otherwise identical to the first study.

The iAUC for TG was not different after intake of LOO containing 14 and 1.8 g lipids compared to placebo and there were no postprandial differences in plasma glucose or insulin levels (data not shown). The iAUC_{0–5 h} for NEFAs was significantly lower after 14 g lipids from LOO compared to placebo (p=0.024) (Fig. 6) and after 3 h the NEFA concentration was significantly lower with 14 g lipids from LOO compared to placebo (p=0.015). There were no differences in plasma total, HDL or LDL cholesterol levels after the different LOO intake and placebo yoghurt (data not shown).

The iAUC_{0–5 h} for PYY and CCK were significantly higher with LOO containing 14 g lipids compared to 1.8 g and placebo (PYY p=0.008 for 14 g vs. 1.8 g and p=0.009 for 14 g vs. placebo, for CCK p=0.007 for 14 g vs. 1.8 g and p=0.002 for 14 g vs. placebo), whereas no significant difference in the iAUC_{0–5 h} for GLP-1 was observed (Fig. 7a–c). Both PYY and CCK were significantly increased with LOO containing 14 g lipids compared to placebo or to LOO containing 1.8 g lipids (PYY p=0.009 and p=0.008 respectively and CCK p=0.001 and p<0.0001, respectively). GLP-1 was higher (p<0.1) or significantly higher after 1 and 5 h with 14 g of LOO compared to placebo (1 h p=0.043, 5 h p=0.064) and there was a trend towards a difference between LOO with 1.8 g lipids and placebo at 3 h (p=0.09).

GLP-2 was only analyzed after a breakfast containing LOO with 14 g lipids and placebo yoghurt. The iAUC_{0–5 h} as well as the concentration of GLP-2 at all postprandial time points was significantly higher after LOO compared to placebo (iAUC_{0–5 h} p=0.002, 1 h p=0.0001, 3 h p=0.04, 5 h p=0.017, and 7 h p=0.029) (Fig. 8).

The participants registered their food intake during the ad libitum served lunch and reminder of that day. Energy intake was 14% lower for women the rest of the day after breakfast with LOO containing 35 g lipids compared to placebo, but this difference was not statistically significant. The average energy intake at lunch was similar around 424 kcal (441 after LOO containing 35 g lipids and 407 kcal after placebo) (range 275–725 in both groups) and not significantly different. The rest of the day was also not different between the three different meals and placebo in the follow-up study (data not shown).

The highest amount of LOO was not very palatable despite the addition of jam. However, LOO containing both 14 and 1.8 g of lipids could easily be mixed with milk powder and yoghurt and flavored; these meals were therefore not very different from placebo.

A few subjects reported an uncomfortable feeling in their gut the first hours after intake of breakfast. However, no difference between any of the test products and placebo could be distinguished.