Energy and macronutrient intake were assessed at two ad libitum meals provided on both days of the exercise and control trials. Specifically, at 2 h a cold buffet meal was provided from which participants were free to consume food ad libitum for 30 min. Acceptability of the buffet items was ensured during the preliminary trial via the completion of a food preference questionnaire (Appendix E). The questionnaire required participants to rate preselected food items on a Likert scale ranging from one (dislike extremely) to ten (like extremely). Participants were excluded from the study if they rated >20% of the items at 4 or less. The items available are listed in Appendix G. The buffet foods were presented identically on each occasion and were in excess of expected consumption. Participants were instructed to eat until ‘comfortably full’ and informed that additional food was available if desired.
At 6 h ad libitum energy intake was assessed using a hot homogenous pasta meal. This method attempted to blind the participants from the amount of food they had eaten. The ad libitum meal was composed of white fusilli (Tesco Fusilli Pasta Twists), tomato sauce (Tesco Bolognese sauce), Cheddar cheese (Tesco British Mature White Cheddar) and olive oil (Tesco Extra Virgin Olive Oil). For each trial, 500 g of pasta was cooked in 2 L of unsalted water at 900 W for 11 min. The pasta was drained with the weight recorded 4 min after cessation of cooking. The cooked pasta was then thoroughly mixed with 156 g of cheese, and 375 g of pasta sauce, until homogenous in nature. The pasta was then covered and allowed to cool completely. Approximately 30 min before the ad libitum meal was due to be served;
50 g of olive oil was added and mixed into the pasta meal. The pasta was then reweighed and
divided equally between 5 smaller serving bowls. Immediately prior to serving the meal, a small bowl was reheated at 900 W for 1 min in the microwave. The bowl of pasta was allowed to cool for 90 s before weighing and presented to the participant. The next bowl was subsequently heated, cooled and replaced the partially consumed bowl. Each returned bowl was weighed. This process was continued until the participant felt ‘comfortably full’, with no time limit set. The macronutrient composition of the meal was balanced (52% carbohydrate, 14% protein and 34% fat), and designed to meet current UK dietary guidelines (Food Standards Agency, 2007).
Participants consumed meals in isolation so that social influence did not affect food selection or quantity eaten (except for the replacement of fresh, and retrieval of used pasta bowls).
Leftovers were weighed and food consumption was determined as the weighed difference of items before and after each meal. The energy and macronutrient content of the items consumed were determined from manufacturers’ values. Previously, authors have shown a high level of reproducibility when measuring energy intake by ad libitum buffets (Arvaniti et al. 2008; CV 8.2%) and hot homogenous meals (Gregersen et al. 2008; CV 14.5%). This suggests that these methods are reliable for assessing ad libitum energy intake in a laboratory environment.
For the time spent away from the laboratory in-between the visits on day one and day two participants were free to select, and consume, a selection of items presented at the cold buffet meal. Participants were permitted to consume these items after leaving the laboratory on day one and until 23:00. The food items available for selection are listed in Appendix H.
Leftovers were returned the next day to determine actual consumption. Water was available ad libitum throughout trials.
4.2.6. Blood sampling
Venous blood samples were collected at 0 and 7 h on day one of each trial, and at 0, 2, 3, 6 and 7 h during day two. Both samples on day one, and the baseline sample on day two, were collected by venepuncture of an antecubital vein. The remaining blood samples on day two (2, 3, 6 and 7 h) were collected using a cannula (Venflon, Becton Dickinson, Helsinborg, Sweden) inserted into an antecubital vein. Patency of the cannula was maintained by flushing with 10 mL of non-heparinised saline (0.9 % (w/v) sodium chloride, Baxter Healthcare Ltd., Norfolk, UK) after each blood sample collection. To avoid dilution of the subsequent sample
(Sarstedt, Leicester, UK) for the measurement of plasma total PYY, acylated ghrelin, leptin, glucose and insulin. To prevent the degradation of acylated ghrelin, blood samples collected into the 4.9 mL monovette were pre-treated and processed following manufacturer instructions (see Chapter 3). The remaining analytes were measured from the resulting plasma collected into the 9 mL monovette. After blood collection, the 9 mL monovette was immediately spun at 1165 x g for 10 min at 4°C in a refrigerated centrifuge (Heraeus Labofuge 400R, Thermo Electron, Osterode, Germany). The plasma supernatant was aliquoted into 2 mL Eppendorfs and stored at -20°C for future analysis.
At each sampling point duplicate 20 µL blood samples were collected into micropipettes for the determination of haemoglobin, and triplicate 20 µL blood samples were collected into heparinised micro haematocrit tubes for the determination of blood haematocrit concentration.
These data were used to estimate plasma volume changes over time (Dill & Costill 1974).
Haematocrit was determined using a microliter-haematocrit centrifuge (MIKRO, 20, Andreas Hettich GmbH and Co.KG, Tuttlingen, Germany). Haemoglobin was determined using cyanmethaemoglobin method with the aid of a spectrophotometer (CECIL CE1011, Cecil Instruments Ltd., Cambridge, UK).
4.2.7. Biochemical analysis
Commercially available enzyme immunoassays were used to determine concentrations of plasma acylated ghrelin (SPI BIO, Montigny le Bretonneux, France), total PYY (Millipore, Billerica, USA), leptin (R&D Systems, Minneapolis, USA) and insulin (Mercodia, Uppsala, Sweden) with the aid of a plate reader to measure absorbance (Expert Plus, ASYS Atlantis, Eugendorf, Austria and Varioskan Flash Multiple Mode Reader, Vantaa, Finland). Plasma glucose concentrations were determined by enzymatic, colorimetric methods using a bench top analyser (Pentra 400; HORIBA ABX Diagnostics, Montpellier, France). Precision of analysis was ensured by quantification of an internal quality control exhibiting high and low values. To eliminate inter-assay variation, samples from each participant were analysed on the same run. The within-batch coefficients of variation for the assays were as follows:
acylated ghrelin 10.9%, total PYY 7.7%, leptin 5.9% insulin 7.9% and glucose 0.6%.