Weight Management and the Prevention of Obesity – The Role of Dietary Protein in the Regulation of Appetite and Food intake

Abstract

Positive energy balance largely driven by an increased motivation to eat is a major cause of body weight (BW) gain in the modern world. Hence, identifying effective dietary interventions to promote satiety and to prevent overconsumption of energy is valuable in preventing obesity. Although dietary protein is more satiating than iso-energetic carbohydrate (CHO) and fat, and holds promise in the suppression of energy intake (EI) for successful BW management, the underlying physiological mechanisms of protein-induced satiety are not fully elucidated. The circulating concentrations of gastrointestinal (GI) peptides, including cholecystokinin (CCK), glucagon-like peptide-1 (GLP-1) and peptide YY (PYY) are commonly proposed and measured as appetite-related biomarkers. The circulating concentrations of amino acids (AAs) are also relevant as possible biomarkers of protein-induced satiety. To advance the science of appetite, the overarching aim of this thesis was to investigate the physiological mechanisms of protein-induced satiety, focusing on the hypothesis of circulating GI peptides and AAs as appetite-related biomarkers. The first study, presented in Chapter 3, aimed to investigate the ‘threshold’ concentration of CCK, GLP-1, and PYY required for successful suppression of motivation to eat. Despite a large body of evidence showing significant postprandial increase in these peptides following a meal, whether this is causal in the suppression of appetite and eating behaviour is not established. Since peptide infusions have been widely demonstrated to successfully suppress hunger and/or EI, this study compared the increase in circulating concentrations of GI peptides observed following dietary preload interventions with concentrations observed following peptide infusion interventions. The biomarker and appetite data compared was obtained from cohorts of lean and healthy adults, previously published. Results showed peak circulating concentrations of GI peptides observed following peptide infusions were greater than dietary preload interventions (Mean fold change above baseline, Infusion vs Diet: CCK, 10.96- vs 2.98-fold; GLP-1, 6.29- vs 1.85-fold; PYY, 4.78- vs 1.82-fold). At least a 3.6-, 4.0- and 3.1-fold increase in CCK, GLP-1 and PYY, respectively, were required to suppress EI following peptide infusion. These respective ‘thresholds’ were achieved through diet in only 29%, 0% and 8% of interventions. Hence, the analysis conducted in this thesis has questioned the physiological role of GI peptides as appetite-related biomarkers in dietary intervention studies. This thesis then proposed that a ‘satiety fingerprint’ comprising circulating concentration of GI peptides plus amino acids (AAs) may better predict appetite responses and EI than individual biomarkers. Hence, in Chapter 4, an Acute Protein Study was conducted to test this hypothesis. The study investigated the dose-response relationship of whey protein ingestion on subjective feelings of appetite, ad lib EI, and GI peptide and AA biomarkers, in a preload challenge design. This study was a randomised controlled 3-arm (control 0 g, low-dose 12.5 g, and high-dose 50 g whey protein; consumed as an iso-volumetric 380 mL beverage) crossover trial, involving a cohort of 24 women with overweight and obesity, aged between 18 – 65 years. Surprisingly, there was no evidence that a higher dose of whey protein slowed the progressive increase in hunger over a 4-hour postprandial period. Furthermore, increasing the dose of whey protein from 0 to 50 g did not suppress ad lib EI when assessed at 4 hours after beverage ingestion, despite a substantial postprandial increase in circulating concentrations of GLP-1, PYY, most proteogenic AAs (except glycine), citrulline and ornithine in a dose-dependent manner. Nevertheless, there were two key novel findings. Firstly, participants consumed less energy when their plasma concentrations of glycine and taurine were higher. Secondly, a ‘satiety fingerprint’ was successfully characterised using a multivariate linear mixed model regression analysis, whereby GLP-1, Gly, Gln, Arg, and Ala emerged as the biomarkers consistently contributed to the estimation of appetite assessed via Visual Analogue Scale (VAS), when analysed with all 3 treatments combined. However, this ‘satiety fingerprint’ explained only ≈20% variance of appetite for an individual participant. The study concluded that the utility of these biomarkers to predict appetite responses over a 4-hour postprandial period following the ingestion of 0 – 50g whey protein beverages is limited in this cohort of women with overweight and obesity, aged between 18 – 65 years . Having shown that an acute protein-induced increase in circulating concentrations of GI peptides and AAs were weak biomarkers of appetite, this thesis then aimed to understand their relationship in a longer-term study. Since energy restriction is known to increase self-reported hunger, whereby a concomitant decrease in circulating concentration of GI peptides has been proposed to be its cause, a higher protein diet is often recommended to promote satiety, leading to a better BW loss outcome. Hence, in Chapter 5, a Weight Loss Study was conducted to manipulate the longer-term circulating concentrations of GI peptides and AAs via a Low Energy Diet (LED) of different macronutrient compositions, and observe the associated change in BW and appetite. This study hypothesised that a higher protein LED may differentially change the circulating GI peptides and AAs profile when compared to a normal protein LED, which in turn associated with the underlying success of a higher protein-induced BW loss. In this study, participants received either Higher Protein (HP) or Normal Protein (NP) treatment, where the concomitant variation in CHO content was accounted for in a 2 × 2 factorial design. A total of 121 women with obesity, aged between 18 – 60 years, were randomised to 1 of 4 treatment groups. Results showed the mean BW loss was 7.9 kg, equivalent to 8.3% initial BW. The HP LED neither enhanced BW loss nor retention of FFM, in comparison to the NP LED. Notably, reported protein intake for all treatments exceeded the Recommended Dietary Intake (RDI) for protein, that is > 0.8 g/kg BW. The reported difference in protein intake between HP and NP LEDs was smaller than anticipated. LC, but not HP, LED was associated with increased postprandial satiety in response to a standardised preload breakfast, but the appetite response could not be explained by the direction of change in biomarkers, in agreement with the findings from the previous chapters. In conclusion, this thesis consistently showed that GLP-1, PYY and AAs were not reliable biomarkers of self-reported appetite, either following acute consumption of whey protein beverages or during longer-term BW loss, in cohorts of young- and middle-aged women with overweight or obesity. Furthermore, a higher protein intake is not always accompanied by a suppression of appetite and/or EI. Hence, public health messages to promote a higher protein intake must be delivered with caution as it clearly must be accompanied by a suppression of EI to benefit BW management. These biomarkers must be very cautiously interpreted by the scientific community when presenting their findings, and they are valuable when presented in combination with VAS and EI data. This thesis also presented the novel use of multivariate linear mixed model regression analysis to characterise a ‘satiety fingerprint’ that can be used to understand the food component or composition that is best related to appetite response.