I am a lecturer in food science in the department of Nutritional Sciences. My research interests are based on plant polysaccharides and their direct and indirect nutritional effects. I originally studied chemistry, but have gradually moved towards food science and nutrition, via plant biology.
Areas of specialism
University roles and responsibilities
- Module Organiser, BMS2053 Food Analysis and Quality Control
- Module Organiser, BMS3059 Food: Chemistry, Functionality and Health Effects
Affiliations and memberships
My research interests are based on plant polysaccharides and their direct and indirect nutritional effects. I am applying this knowledge to a wide range of projects involving industrial and academic partners.
Postgraduate research supervision
Design of functional oligosaccharides as sucrose replacers
Postgraduate research supervision
Resistant starch 5: amylose-lipid complexes
I have taught food science to undergraduate students for the last 6 years, first at King's College London and now at the University of Surrey. I supervise undergraduate and postgraduate projects related to food science, often involving new product development. I am an Associate Fellow of the Higher Education Academy. To see me during office hours, please email me for an appointment.
Courses I teach on
Almonds contain phytochemicals and nutrients that potentially have positive health benefits in relation to heart disease, diabetes and obesity. One important mechanism associated with these benefits is the bioaccessibility (release) of lipids from the almond tissue during mastication and digestion. Indeed, it has been reported that in human subjects on almond-rich diets, a significant proportion of lipid remains undigested, mainly caused by the entrapment of lipid by intact cell walls (dietary fiber). However, the bioaccessibility of almond lipid has not yet been quantified. We have studied this by performing a mastication trial in healthy human volunteers (n = 15). The particle size distribution of chewed almonds was measured using mechanical sieving and laser diffraction methods. Lipid release from ruptured almond cells was estimated using a theoretical model based on the particle size of the chewed almond tissue and the diameter of the lipid-rich cells. Our results showed that laser diffraction was the most reliable and efficient method. Lipid release from masticated almonds was estimated to be ~30% of total lipid and originated from the ruptured cells of the fractured tissue. This indicates that the majority of almond lipid is unavailable for the early stages of digestion. Further work will extend to studies of lipid release and digestion at more distal sites of the gastrointestinal tract
Ingestion of different foods containing identical amounts of starch can result in very different postprandial rises in blood glucose and insulin concentrations. Limitation of the early rises in blood glucose and insulin levels seems to be beneficial to human health in the long term. Many studies of starch digestion in vitro are made to understand the molecular basis for differences in digestion rates in vivo to enable prediction of likely rates of digestion of particular starchy foods. Michaelis-Menten kinetics of starch digestibility provides estimates of available (digestible) substrate as starch samples are hydrothermally treated. Combined with studies of starch structure using calorimetry and FTIR spectroscopy, key features influencing rates of amylolysis were identified. Measurement of product formation during prolonged incubations of starch with α-amylase produces digestibility curves. Use of logarithm of slope (LOS) plots to analyse the curves by 1st order kinetics gave values for digestibility rate constants and the total digestible starch, C∞. An important conclusion is that contrary to many reports in the literature, cooked starches do not contain distinct fractions of rapidly and slowly digested material. These kinetic approaches have also been used in studies of plan-tencapsulated starch to understand how cell walls influence access of amylase to starch.
Seven project update posters
Series of five project update presentations
Polysaccharides derived from plant foods are major components of the human diet, with limited contributions of related components from fungal and algal sources. In particular, starch and other storage carbohydrates are the major sources of energy in all diets, while cell wall polysaccharides are the major components of dietary fibre. We review the role of these components in the human diet, including their structure and distribution, their modification during food processing and effects on functional properties, their behaviour in the gastro-intestinal tract and their contribution to healthy diets.
Abstract 149/1077 (Poster)
We have previously reported on the low lipid bioaccessibility from almond seeds during digestion in the upper gastrointestinal tract (GIT). In the present study, we quantified the lipid released during artificial mastication from four almond meals: natural raw almonds (NA), roasted almonds (RA), roasted diced almonds (DA) and almond butter from roasted almonds (AB). Lipid release after mastication (8.9% from NA, 11.8% from RA, 12.4% from DA and 6.2% from AB) was used to validate our theoretical mathematical model of lipid bioaccessibility. The total lipid potentially available for digestion in AB was 94.0%, which included the freely available lipid resulting from the initial sample processing and the further small amount of lipid released from the intact almond particles during mastication. Particle size distributions measured after mastication in NA, RA and DA showed most of the particles had a size of 1000 µm and above, whereas AB bolus mainly contained small particles (
Introduction The proportion of lipid released during mastication of nuts is strongly influenced by particle size, due to natural encapsulation of the lipid by the walls of intact cells. The cell walls may act as barriers to digestion and may partially explain why nuts have reduced metabolizable energy versus the energy content predicted by Atwater factors. The lipid released from masticated nuts can be calculated using a mathematical model which has the cell diameter and the particle size distribution (PSD) of the bolus as variables. This study measured the cell size and PSD of four tree nuts (raw cashews, raw walnuts, roasted pistachios and raw Brazil nuts) to predict the lipid released due to their mastication. Objectives To predict the proportion of lipid released from masticated tree nuts, using measurements of cell size and PSD of masticated tree nuts (including cashews, walnuts, pistachios and Brazil nuts). To compare the results for these nuts to those already published for almonds. Method/Design Transverse and longitudinal sections were cut from each nut (including almonds) and micrographs processed using image analysis software to calculate the average cell diameter. Two randomized, un-blinded, cross-over trials were conducted. In each trial, 10 healthy volunteers attended two sessions, at which they chewed eight samples of a randomly allocated nut. For determination of PSDs, the expectorated boluses were sieved (2 boluses) or analyzed by laser diffraction (2 boluses). Initial lipid release was then predicted using the mathematical model. Results The cashew cells (34.3 μm) were smaller than the almond (45.1 μm), walnut (49.4 μm), pistachio (53.1 μm), and Brazil nut cells (60.8 μm). Laser diffraction showed that masticated nut boluses had median particle sizes (± SEM) which were smaller (cashews, 178 ± 12 μm; walnuts, 179 ± 8 μm; pistachios, 123 ± 10 μm; Brazil nuts, 145 ± 8 μm) than that for almonds measured previously (550 ± 18 μm). This results in higher predicted lipid release, calculated from the mathematical model, (mean, range) for cashews (12.3%, 8.7–16.3%), walnuts (14.5%, 12.0–18.0%), pistachios (11.0%, 8.7–13.0%) and Brazil nuts (14.4%, 11.9–17.3%) than for almonds (9.5%, 7.4–11.1%). Conclusions All of the five nuts had predicted lipid releases of less than 18%, which would be expected to attenuate postprandial lipemia and total nutrient availability relative to that expected due to their total lipid content. Due to their higher lipid content and predicted lipid release, walnuts and Brazil nuts (after mastication) are likely to release more fat on mastication than the other nuts tested.
BACKGROUND: Cereal crops, particularly wheat, are a major dietary source of starch, and the bioaccessibility of starch has implications for postprandial glycemia. The structure and properties of plant foods have been identified as critical factors in influencing nutrient bioaccessibility; however, the physical and biochemical disassembly of cereal food during digestion has not been widely studied. OBJECTIVES: The aims of this study were to compare the effects of 2 porridge meals prepared from wheat endosperm with different degrees of starch bioaccessibility on postprandial metabolism (e.g., glycemia) and to gain insight into the structural and biochemical breakdown of the test meals during gastroileal transit. DESIGN: A randomized crossover trial in 9 healthy ileostomy participants was designed to compare the effects of 55 g starch, provided as coarse (2-mm particles) or smooth (
BACKGROUND: The particle size and structure of masticated almonds have a significant impact on nutrient release (bioaccessibility) and digestion kinetics. OBJECTIVES: The goals of this study were to quantify the effects of mastication on the bioaccessibility of intracellular lipid of almond tissue and examine microstructural characteristics of masticated almonds. DESIGN: In a randomized, subject-blind, crossover trial, 17 healthy subjects chewed natural almonds (NAs) or roasted almonds (RAs) in 4 separate mastication sessions. Particle size distributions (PSDs) of the expectorated boluses were measured by using mechanical sieving and laser diffraction (primary outcome). The microstructure of masticated almonds, including the structural integrity of the cell walls (i.e., dietary fiber), was examined with microscopy. Lipid bioaccessibility was predicted by using a theoretical model, based on almond particle size and cell dimensions, and then compared with empirically derived release data. RESULTS: Intersubject variations (n = 15; 2 subjects withdrew) in PSDs of both NA and RA samples were small (e.g., laser diffraction; CV: 12% and 9%, respectively). Significant differences in PSDs were found between these 2 almond forms (P < 0.05). A small proportion of lipid was released from ruptured cells on fractured surfaces of masticated particles, as predicted by using the mathematical model (8.5% and 11.3% for NAs and RAs, respectively). This low percentage of lipid bioaccessibility is attributable to the high proportion (35-40%) of large particles (>500 μm) in masticated almonds. Microstructural examination of the almonds indicated that most intracellular lipid remained undisturbed in intact cells after mastication. No adverse events were recorded. CONCLUSIONS: Following mastication, most of the almond cells remained intact with lipid encapsulated by cell walls. Thus, most of the lipid in masticated almonds is not immediately bioaccessible and remains unavailable for early stages of digestion. The lipid encapsulation mechanism provides a convincing explanation for why almonds have a low metabolizable energy content and an attenuated impact on postprandial lipemia.
food matrix (muffins) investigating whether the cell-wall barrier regulates the bioaccessibility of nutrients within this matrix. Muffins containing small (AF) or large (AP) particles of almond were digested in triplicate using an in vitro dynamic gastric model (DGM, 1 h) followed by a static duodenal digestion (8 h). AF muffins had 97.1 ± 1.7% of their lipid digested, whereas AP muffins had 57.6 ± 1.1% digested. In vivo digestion of these muffins by an ileostomy volunteer (0–10 h) gave similar results with 96.5% and 56.5% lipid digested, respectively. The AF muffins produced a higher postprandial triacylglycerol iAUC response (by 61%) than the AP muffins. Microstructural analysis showed that some lipid remained encapsulated within the plant tissue throughout digestion. The cell-wall barrier mechanism is the main factor in regulating lipid bioaccessibility from almond particles.
Processed foods are typically praised/revered for their convenience, palatability, and novelty; however, their healthfulness has increasingly come under scrutiny. Classification systems that categorise foods according to their “level of processing” have been used to predict diet quality and health outcomes and inform dietary guidelines and product development. However, the classification criteria used are ambiguous, inconsistent and often give less weight to existing scientific evidence on nutrition and food processing effects; critical analysis of these criteria creates conflict amongst researchers. We examine the underlying basis of food classification systems and provide a critical analysis of their purpose, scientific basis, and distinguishing features by thematic analysis of the category definitions. These classification systems were mostly created to study the relationship between industrial products and health. There is no consensus on what factors determine the level of food processing. We identified four defining themes underlying the classification systems: 1. Extent of change (from natural state); 2. Nature of change (properties, adding ingredients); 3. Place of processing (where/by whom); and 4. Purpose of processing (why, essential/cosmetic). The classification systems embody socio-cultural elements and subjective terms, including home cooking and naturalness. Hence, “processing” is a chaotic conception, not only concerned with technical processes. Most classification systems do not include quantitative measures but, instead, imply correlation between “processing” and nutrition. The concept of “whole food” and the role of the food matrix in relation to healthy diets needs further clarification; the risk assessment/management of food additives also needs debate. •Classifications were created to link processed foods to nutrition and health outcomes.•There is no consensus on what determines the level of food processing.•Classification includes the extent, nature of change, place and purpose of processing.•Processed food concepts relate to naturalness, additives, convenience, home cooking.•Food classifications embody social and cultural elements and subjective terms.
Background: Processed foods are typically praised/revered for their convenience, palatability, and novelty; however, their healthfulness has increasingly come under scrutiny. Classification systems that categorise foods according to their “level of processing” have been used to predict diet quality and health outcomes and inform dietary guidelines and product development. However, the classification criteria used are ambiguous, inconsistent and often give less weight to existing scientific evidence on nutrition and food processing effects; critical analysis of these criteria creates conflict amongst researchers. Scope and approach: We examine the underlying basis of food classification systems and provide a critical analysis of their purpose, scientific basis, and distinguishing features by thematic analysis of the category definitions. Key findings and conclusions: These classification systems were mostly created to study the relationship between industrial products and health. There is no consensus on what factors determine the level of food processing. We identified four defining themes underlying the classification systems: 1. Extent of change (from natural state); 2. Nature of change (properties, adding ingredients); 3. Place of processing (where/by whom); and 4. Purpose of processing (why, essential/cosmetic). The classification systems embody socio-cultural elements and subjective terms, including home cooking and naturalness. Hence, “processing” is a chaotic conception, not only concerned with technical processes. Most classification systems do not include quantitative measures but, instead, imply correlation between “processing” and nutrition. The concept of “whole food” and the role of the food matrix in relation to healthy diets needs further clarification; the risk assessment/management of food additives also needs debate.
Although the growth of bacteria has been studied for more than a century, it is only in recent decades that surface-associated growth has received attention. In addition to the well-characterized biofilm and swarming lifestyles, bacteria can also develop as micro-colonies supported by structured environments in both food products and the GI tract. This immobilized mode of growth has not been widely studied. To develop our understanding of the effects of immobilization upon a food-borne bacterial pathogen, we used the IFR Gel Cassette model. The transcriptional programme and metabolomic profile of Salmonella enterica serovar Typhimurium ST4/74 were compared during planktonic and immobilized growth, and a number of immobilization-specific characteristics were identified. Immobilized S.Typhimurium did not express motility and chemotaxis genes, and electron microscopy revealed the absence of flagella. The expression of RpoS-dependent genes and the level of RpoS protein were increased in immobilized bacteria, compared with planktonic growth. Immobilized growth prevented the induction of SPI1, SPI4 and SPI5 gene expression, likely mediated by the FliZ transcriptional regulator. Using an epithelial cell-based assay, we showed that immobilized S.Typhimurium was significantly less invasive than planktonic bacteria, and we suggest that S.Typhimurium grown in immobilized environments are less virulent than planktonic bacteria. Our findings identify immobilization as a third type of surface-associated growth that is distinct from the biofilm and swarming lifestyles of Salmonella.
Dietary fibre is a generic term for a chemically diverse group of carbohydrates that are resistant to endogenous enzymes of the human digestive tract. The major component of dietary fibre consists of plant cell walls, which are supramolecular structures, composed of complex heterogeneous networks of cellulose, hemicelluloses and pectic substances. The amounts and relative proportions of these carbohydrates vary depending on the type and maturity of the plant tissue. Some plant cell walls, especially those from leguminous seeds, are rich in water-soluble non-starch polysaccharides (NSP). Intact plant cell walls and NSP affect the rate and extent of nutrient digestion, with important implications for health and disease. Certain types of fibre reduce the rate of starch digestion, which in turn can significantly attenuate the postprandial rise in blood glucose and insulin concentrations. This is potentially beneficial in the prevention and treatment of diseases, including diabetes mellitus and cardiovascular disease. However, the mechanisms of action of NSP in relation to the digestive process are still not well understood. They are thought to include formation of viscous solutions, encapsulation of nutrients and inhibition of digestive enzymes. These mechanisms are illustrated using specific examples. Oat β-glucan is used to show the effects of various processing techniques on β-glucan molecular weight, and hence viscosity, on risk factors for diabetes and cardiovascular disease. Evidence for guar galactomannan acting as an inhibitor of α-amylase, in addition to forming viscous solutions, is presented. Finally, the effect of intact plant cell walls on the bioaccessibility of nutrients is discussed.
Investigators often study product release from starches during prolonged incubations with α-amylase in vitro. The reaction time courses usually fit to a linear form of a first order rate equation, i.e., ln[(C∞ − Ct)/C∞] = −kt. This equation calls for an accurate estimate of C∞, i.e., the concentration of product at the end of the reaction. Estimates of C∞ from digestibility curves can be unreliable. The Guggenheim method does not require prior knowledge of C∞ but seems not to have been applied to starch hydrolysis data. An alternative method is also available in which the logarithm of the slope (LOS) of a digestibility curve at various time points is plotted against time. This allows estimations of both k and C∞ and can also reveal whether changes occur in digestion rate from rapid to slow as digestion proceeds. We describe the Guggenheim and LOS methods and provide examples of their application to starch digestibility data.
Chinese water chestnut (Eleocharis dulcis (Burman f.) Trin ex Henschel) is a corm consumed globally in Oriental-style cuisine. The corm consists of three main tissues, the epidermis, subepidermis, and parenchyma; the cell walls of which were analyzed for sugar, phenolic, and lignin content. Sugar content, measured by gas chromatography, was higher in the parenchyma cell walls (931 μg/mg) than in the subepidermis (775 μg/mg) or epidermis (685 μg/mg). The alkali-extractable phenolic content, measured by high-performance liquid chromatography, was greater in the epidermal (32.4 μg/mg) and subepidermal cell walls (21.7 μg/mg) than in the cell walls of the parenchyma (12.3 μg/mg). The proportion of diferulic acids was higher in the parenchyma. The Klason lignin content of epidermal and subepidermal cell walls was ~15%. Methylation analysis of Chinese water chestnut cell-wall polysaccharides identified xyloglucan as the predominant hemicellulose in the parenchyma for the first time, and also a significant pectin component, similar to other nongraminaceous monocots.
A number of studies have demonstrated that consuming almonds increases satiety but does not result in weight gain, despite their high energy and lipid content. To understand the mechanism of almond digestion, in the present study, we investigated the bioaccessibility of lipids from masticated almonds during in vitro simulated human digestion, and determined the associated changes in cell-wall composition and cellular microstructure. The influence of processing on lipid release was assessed by using natural raw almonds (NA) and roasted almonds (RA). Masticated samples from four healthy adults (two females, two males) were exposed to a dynamic gastric model of digestion followed by simulated duodenal digestion. Between 7·8 and 11·1 % of the total lipid was released as a result of mastication, with no significant differences between the NA and RA samples. Significant digestion occurred during the in vitro gastric phase (16·4 and 15·9 %) and the in vitro duodenal phase (32·2 and 32·7 %) for the NA and RA samples, respectively. Roasting produced a smaller average particle size distribution post-mastication; however, this was not significant in terms of lipid release. Light microscopy showed major changes that occurred in the distribution of lipid in all cells after the roasting process. Further changes were observed in the surface cells of almond fragments and in fractured cells after exposure to the duodenal environment. Almond cell walls prevented lipid release from intact cells, providing a mechanism for incomplete nutrient absorption in the gut. The composition of almond cell walls was not affected by processing or simulated digestion.