Food Research International 88 (2016) 181190

Contents lists available at ScienceDirect

Food Research International

journal homepage: www.elsevier.com/locate/foodres

Modelling the breakdown mechanics of solid foods during

gastric digestion
Krista C. Drechsler, Maria J. Ferrua
Riddet Institute, Massey University, Private Bag 11 222, Palmerston North 4442, New Zealand

a r t i c l e

i n f o

Article history:
Received 11 August 2015
Received in revised form 23 February 2016
Accepted 26 February 2016
Available online 3 March 2016
Keywords:
Gastric digestion
Food material properties
Breakdown mechanics
Solid foods

a b s t r a c t
Solid food disintegration within the stomach has a major role on the rate and nal bioavailability of nutrients
within the body. Understanding the link between food material properties and their behaviour during gastric digestion is key to the design of novel structures with enhanced functionalities. However, despite extensive research, the establishment of proper relationships has proved difcult. This work builds on the hypothesis that
to bridge this knowledge gap a better understanding of the underlying mechanisms of food disintegration during
digestion is needed. The purpose of this study is to propose a new protocol that, by uncoupling the physicochemical processes occurring during gastric digestion, allows for a more rigorous understanding of these mechanisms.
Using steamed potatoes as a product model, this study aims to develop a viable methodology to characterize the
role of gastric juice and compressive forces on the breakdown mechanics of solid foods during digestion. From a
general viewpoint, this work not only reveals the importance of the parameter used to describe the size distribution of food particles on the interpretation of their breakdown behaviour, but also provides a new framework to
characterize the mechanisms involved. Results also illustrate that food breakdown during gastric digestion might
well not follow a unimodal behaviour, highlighting the need to characterize their performance based on parameters describing broad aspects of their particle size distribution rather than single point values. Arguably simplistic on its approach, this study illustrates how an improved understanding of the role of chemical and physical
processes on the breakdown mechanics of solid foods can facilitate valid inferences with respect to their invivo performance during digestion. In particular, it shows that while the contraction forces occurring in the stomach can easily disintegrate the potato matrix at the molecular level, the continuous exposure to gastric juices will
promote their disintegration into progressively smaller debris. A discussion on the challenges and future directions for the implementation of a more general and standardized protocol is provided. Not intended to reproduce
the breakdown behaviour of foods during gastric digestion, but rather to characterize the mechanisms involved,
the proposed protocol would open new opportunities to identify the material properties governing the performance of different foods upon ingestion.
2016 Elsevier Ltd. All rights reserved.

1. Introduction
The role of food structures on the nutritional and functional performance of foods during digestion is by now widely recognized in the
area of food research (McClements & Li, 2010; Nicolai & Durand, 2013;
Norton, Wallis, Spyropoulos, Lillford, & Norton, 2014; Parada &
Aguilera, 2007; Turgeon & Rioux, 2011; van Aken, 2010; Van Kleef,
Van Trijp, Van Den Borne, & Zondervan, 2012).
Upon ingestion, foods undergo a number of physicochemical changes
that lead to their disintegration, transport, and absorption into the body.
While the result of a complex series of physicochemical processes occurring along the entire gastrointestinal tract, research has evidenced the
signicant role of gastric digestion on the rate and nal bioavailability
Corresponding author.
E-mail addresses: kcdrechsler@gmail.com (K.C. Drechsler), m.j.ferrua@massey.ac.nz
(M.J. Ferrua).

http://dx.doi.org/10.1016/j.foodres.2016.02.019
0963-9969/ 2016 Elsevier Ltd. All rights reserved.

of nutrients (Le Feunteun et al., 2014; Mackie, Raee, Malcolm, Salt, &
van Aken, 2013; Marciani et al., 2007, 2012, 2013).
Second only to the mouth, the stomach is the main site for food disintegration during digestion. In the case of solid foods, their breakdown
behaviour within the stomach is expected to have a major contribution
to digestion outcomes (Kong & Singh, 2008). The higher their level of
breakdown, the higher their exposure to gastric juices, and their subsequent level of chemical degradation and nutrient release during the process. In addition, since the stomach retains particles larger than 3 to
4 mm (Mllertz, 2010), food breakdown within the stomach will not
only inuence the bioaccessibility of nutrients reaching the small intestine, but also their rate of delivery into the organ (Urbain et al., 1989).
The inuence of gastric disintegration and transit time on the absorption and plasma prole of solid drug tables is well established within the pharmaceutical area (Abrahamsson, Albery, Eriksson, Gustafsson,
& Sjberg, 2004; Brouwers, Tack, & Augustijns, 2007; Kelly et al., 2003;
Koziolek et al., 2014; Walter-Sack, de Vries, Nickel, Stenzhorn, &

182

K.C. Drechsler, M.J. Ferrua / Food Research International 88 (2016) 181190

Weber, 1989). Within the food sector, a growing body of scientic evidence illustrates the link that exists between the disintegration of
solid foods in the stomach, the rate of gastric emptying and the nal bioavailability of nutrients within the body (Bornhorst, Ferrua, Rutherfurd,
Heldman, & Singh, 2013; Lamothe, Corbeil, Turgeon, & Britten, 2012;
Guo et al., 2015; Guo, Ye, Lad, Dalgleish, & Singh, 2016).
Over the past decade, scientists have strived to understand how food
material properties inuence their performance during digestion
(Bornhorst & Singh, 2014); knowledge essential to assist the design of
novel structures with enhanced functionalities. However, despite significant advances in knowledge, the establishment of proper relationships
has proven difcult.
Depending on the food, when gastric juice penetrates the matrix, a
number of structural changes start to occur. By modifying the material
properties of the ingested food, these structural changes will modulate
its mechanical disintegration when exposed to the peristaltic and
grinding activity of the stomach. Only a better understanding of the
mechanisms underlying these processes will allow for a better prediction of the performance of different formulations and structures during
digestion.
Arguably realistic in approach, current in-vivo and in-vitro studies
can only report on the response of foods to the complex and dynamic
synergy of physicochemical processes involved. As such, their ability
to inform on those key material properties governing the in-vivo performance of foods during digestion is largely limited.
The purpose of this study is to propose a new protocol that, by
uncoupling the physicochemical processes occurring during gastric digestion, allows for a better understanding of the mechanisms of solid
food disintegration during the process. In particular, using potatoes as
a product model, this study aims to develop a viable methodology to
characterize the mechanisms and extent of particle breakdown, in response to the different physicochemical processes involved.
2. Materials and methods
2.1. Chemicals and materials
Nadine potato tubers were purchased from a local store and used
within no more than three days. Potatoes were stored under dark, ventilated, cool conditions (15 C) inside a perforated plastic bag.
Pepsin from porcine gastric mucosa (0.7 FIP unit/mg) was purchased
from Merck (Darmstadt, Germany). Mucin from porcine stomach (Type
III, bound sialic acid 0.51.5%, partially puried powder) was purchased
from Sigma-Aldrich (St Louis, MO, USA). HCl was purchased from EMD
Millipore Corporation (Billerica, MA, USA). All other chemicals were of
analytical grade.
Simulated gastric uid (SGF) of pH 1.8 (3 gL1 pepsin, 1.5 gL1
mucin, 8.775 gL 1 NaCl) was prepared (Guo, Ye, Lad, Dalgleish, &
Singh, 2014) using ultrapure water (Milli-Q water system, Millipore
Corporation, MA, USA).
2.2. Food sample preparation
Potato cubes (5 mm side length) were cut using a ULTRA 2.0
MASTER Mandolin (Le Buyer, France).
Potato cubes were steamed using a steamer pot (25 cm diameter)
whose perforated plate was substituted by a 1.5 mm stainless steel
sieve mesh. One layer of potato cubes (250275 g) was steamed over
1 L of boiling water for 4 min, removed and immersed into 5 L of cold
water to limit further cooking.

maintained at 37 C within a static water bath (HAAKE SC100, Thermo

Scientic). Potatoes were soaked for up to 4 h (0, 1, 2, 3 and 4 h).
In an attempt to standardize the protocol and reduce the inuence of
the buffering capacity of foods on the outcome of the process, potatoes
were rst soaked in 650 mL of acidic water (pH 1.8) for 5 min. The acidic
water was discarded and the potatoes were then soaked in SGF. The pH
of the SGF was readjusted every 20 min to 1.8 using HCl (1N), with no
more than two pH adjustments been required during the process. The
soaking process was considered to start once the samples were immersed in SGF.
After soaking, the potato cubes were removed from the unit and carefully rinsed with running water to remove the excess of SGF from their
surfaces. Soaked samples were immediately analysed for their mechanical properties and breakdown behaviour (as described subsequently).
2.4. Mechanical properties
The mechanical properties of the potato samples before and after
soaking in SGF were monitored using a texture analyser TA-XTPlus (Stable Microsystems, Surrey, UK). The strength, apparent elasticity, and
toughness of the potato matrix were determined using a uni-axial compression test, while the surface hardness of the cubes was determined
using a modied protocol of the Brinell hardness method.
2.4.1. Single compression test
During compression, two potato cubes were compressed to a strain
of 60% using a cylindrical probe (4 cm diameter). The test was performed at a speed of 0.3 mms1 and a trigger force of 0.09 N. The failure strength (fracture force), apparent elasticity, and toughness of the
samples were determined as illustrated in Fig. 1.
In particular, their apparent elasticity was determined as the slope
of the stressstrain curve between one and two-thirds of the point of
fracture (Canet, Alvarez, & Gil, 2007).
2.4.2. Modied Brinell surface hardness test
The surface hardness of the potato matrix was determined by measuring how much a 2 mm stainless steel sphere indented the potato surface when compressed against it at a constant force of 0.15 N for a 10 s
period.
The applied force and dwell were selected to prevent the sphere
from penetrating more than 1 mm inside the potato, regardless of the
level of soaking investigated (Fig. 1). A hardness number (HN) was
then computed as follows (Instron, 2015):

HN

2F
p

2
D D D2 d

where F is the applied force (N), D is the diameter of the sphere (mm)
and d is the diameter of the indentation (mm). Under the scope of
this study, the diameter of the indentation was computed based on
the diameter and penetration distance of the sphere.
The inuence of soaking time on any of the material properties previously discussed was investigated in triplicate, with at least 9 subsamples being analysed per trial.
2.5. Breakdown behaviour (extent and mechanisms)
The inuence of gastric juice on the breakdown behaviour of the potato cubes was analysed by exposing the soaked samples to different,
well-dened bulk compression conditions.

2.3. Soaking in SGF (static digestion)

A monolayer of steamed potato cubes (250275 g) was suspended
in 650 mL of SGF (2.4 L microwave steamer, Systema Plastics) and

2.5.1. Bulk compression

A bed of potato cubes (45 g potato, 17 mL water) was placed inside a
cylindrical cell (4.2 cm diameter) and compressed at a speed of

K.C. Drechsler, M.J. Ferrua / Food Research International 88 (2016) 181190

183

Fig. 1. Experimental analysis of the mechanical properties and breakdown behaviour of the potato matrix.

2 mms1 Fig. 1). The effect of four different compression levels was investigated (20, 40, 60 and 80% strain).
The physical breakdown of foods within the stomach is largely associated with the compressive activity of a series of periodic and synchronized contraction of the gastric wall. The mechanical stresses generated
as a result of these contractions have been investigated in-vivo and
found to be highly dependent on the physical properties of the digesta.
Mechanical stresses of 5.1 kPa and 82.2 kPa were reported in association
with the ingestion of a water like uid (Marciani et al., 2001) and 220 g
of cooked rice with 350 mL of tea (Kamba et al., 2000), respectively.
Values of up to 150 kPa were reported in dog stomachs after the

consumption of 200 g of a solid meal (Purina, Japan) with only 20 mL

of water.
The bulk compression test used in this study was not intended to reproduce the physical stresses that develop during digestion, but to provide an initial framework to characterize the inuence of gastric juice on
the breakdown mechanics of the potato samples. Nevertheless, the mechanical stresses recorded during the trials were largely within the
range recorded in-vivo (5.9 0.7 kPa to 129 20 kPa), except for samples soaked for less than 2 h and compressed to levels higher than 60%.
The highest level of mechanical stresses measured during the trials
(384 20 kPa) was associated with a bed of steamed potatoes

Table 1
Mixed Weibull distribution parameters. Experimental averages (triplicate) STD.
Soaking time (h)

Compression (%)

1 (103)

k1

k2

r2

0
0
0
1
1
1
1
2
2
2
2
3
3
3
3
4
4
4
4

20
40
60
20
40
60
80
20
40
60
80
20
40
60
80
20
40
60
80

0.26 0.18
0.34 0.08
0.39 0.10
0.09 0.04
0.18 0.04
0.44 0.03
0.51 0.04
0.14 0.02
0.29 0.06
0.46 0.07
0.55 0.04
0.23 0.09
0.37 0.14
0.56 0.06
0.58 0.03
0.19 0.09
0.42 0.03
0.58 0.08
0.65 0.04

1.16 0.65
1.45 0.26
4.66 0.68
1.66 0.68
4.22 2.39
7.57 3.28
5.13 1.59
1.50 0.45
4.81 1.14
6.01 0.79
4.33 0.67
0.94 0.36
3.48 2.39
5.46 0.71
4.59 0.55
3.16 0.26
3.97 0.75
4.05 0.75
3.36 0.38

0.98 0.25
0.87 0.13
0.84 0.06
0.84 0.09
0.82 0.11
0.98 0.17
1.26 0.10
0.87 0.09
0.89 0.11
1.03 0.02
1.27 0.12
1.02 0.12
0.81 0.04
0.87 0.06
1.09 0.15
1.24 0.08
1.21 0.04
1.18 0.16
1.33 0.14

1.76 0.36
1.80 0.19
1.74 0.29
1.94 0.21
1.92 0.30
1.80 0.11
0.93 0.08
1.72 0.16
1.84 0.26
1.80 0.23
0.90 0.37
1.82 0.05
2.00 0.05
2.00 0.07
0.76 0.46
2.02 0.15
2.14 0.02
1.78 0.65
1.07 0.62

7.09 1.90
5.87 1.23
2.76 0.80
6.07 0.84
5.38 0.80
1.80 0.81
0.69 0.07
5.72 0.34
4.34 0.74
1.67 0.69
0.66 0.13
7.13 1.54
4.73 2.00
1.47 0.35
0.63 0.17
5.60 0.94
4.48 1.32
1.48 0.82
0.64 0.12

N0.996
N0.996
N0.995
N0.996
N0.9982
N0.993
N0.997
N0.997
N0.996
N0.994
N0.997
N0.998
N0.996
N0.994
N0.997
N0.998
N0.996
N0.994
N0.996

184

K.C. Drechsler, M.J. Ferrua / Food Research International 88 (2016) 181190

compressed to a 60% level (max level of compression considered for

non-soaked samples in this study).
The inuence of soaking time and compression level on the breakdown mechanics of the potatoes was investigated in triplicate.
2.5.2. Particle size distribution image analysis
The breakdown mechanics of the compressed samples was characterized based on an image analysis of their particle size distributions
(PSD).
2.5.2.1. Sampling and image acquisition. The rst step of this analysis was
to dene the minimum amount of sample that needs to be imaged to accurately reproduce the particle size distribution of the compressed bed.
After a series of preliminary tests, a sample size of 5 mL was selected.
Three (5 mL) subsamples were randomly taken for each compressed
bed and distributed among as many rectangular petri dishes
(12.2 cm 8.0 cm) as needed to avoid particles touching each other during imaging. The higher the soaking time and compression level analysed,
the larger the extent of particle breakdown, and the larger the number of
petri dishes required (about 3 to 8).
Particles were dyed with an iodine solution (Lugol's 1%) to increase
their contrast against the background. About 2030 mL of ultrapure
water was added to each petri dish to help suspend and separate the
particles. Each petri dish was individually placed on top of a light box
and imaged (Nikon 1 V1, AF-S DX Micro NIKKOR 40 mm f/2.8G
macrolens).
2.5.2.2. Numerical algorithm. The size distribution of the compressed particles was determined using Matlab 2013a (MathWorks, Natick, Mass.,
USA). A custom built code was written to determine the number and
surface area of all the particles present in each individual petri dish associated with the same 5 mL subsample.
Images from each petri dish were rst converted to a grayscale
image. The contrast between particles and background was enhanced
by readjusting the intensity range between the lowest 1% and highest
2% of the data. An intensity threshold that differentiated particles from
the background was determined by averaging the intensity of eight
manually-selected pixels that represent the lighter potato particles
found in a magnied ( 2.5) representation of the image. Using this
threshold, the image was converted to black and white, with black regions representing particles. The borders of the petri dish were removed
and all interconnected black pixels identied as an individual component (particle). Each particle was labelled and its projected area
computed and stored for further analysis. The code developed to determine the particle size distribution of the compressed samples is published online along the electronic version of this article.
The size distribution of the particles (in terms of their relative numbers and associated surface areas) was computed and averaged over all
subsamples. To facilitate the comparison among different treatments,
the size distribution of the particles was always analysed over one hundred size classes ranging from 104 mm2 to 100 mm2.
2.5.3. Breakdown model
When using image analysis, the breakdown behaviour of chewed/
digested food has been described in terms of the number of particles
of different sizes present in the sample (Bornhorst, Kostlan, & Singh,
2013; Hutchings et al., 2012; Jalabert-Malbos, Mishellany-Dutour,
Woda, & Peyron, 2007; Mishellany, Woda, Labas, & Peyron, 2006). In
particular, it is a common practice to t the cumulative distribution of
particles in terms of either their numbers or surface areas to a Rosin
Rammler model (unimodal Weibull distribution), and quantify their extent of breakdown based on its median and spread values.
While fragmentation and surface erosion have been proposed as
mechanisms for the disintegration of solid foods in the stomach (Kong
& Singh, 2009), no methodology has been yet established to identify
their contribution to the process. As part of this study, a framework

commonly used in the area of powder technologies was implemented

to characterize the damage mechanisms undergone by the potato samples during compression. As illustrated in Fig. 2, based on the size distribution of compressed particles in relationship to their initial size, it is
possible to quantify the amount of product associated with undamaged
particles, fragments and ne debris (Ghadiri, Ning, Kenter, & Puik,
2000). This information can be then be used to inform on the relative
relevance of the different damage mechanisms involved (see
Section 3.2.2).
Unlike previous work, a RosinRammler distribution could not t
the breakdown behaviour of the compressed samples in this study.
Characterized by a bimodal behaviour, the distribution of product
among different particle sizes was modelled using a mixed Weibull distribution function (Eq. (2)).
0

 k1 1

y @1e

x
1

 k2 1

A 1@1e

x
2

where y is the cumulative particle size distribution of the sample (0 to

1), x is the particle area in relationship to its initial value (A/A0), is a
mixing weight parameter that represents the proportion of the rst
mode, and 1, k1, 2, and k2 are the scale and shape parameters describing the distribution of each mode respectively.
The selection of a mixed Weibull model was based on its ability to describe the breakdown behaviour of a wide range of particulate systems,
including the breakdown of gold nanoparticles by pulsed laser irradiation,
of cotton bres during processing, and of peanut particles by oral mastication (Krifa, 2009; Kim, Morgenstern, Bronlund, Fosterd, & Le Gote, 2011).
The experimental data was tted to the model using a non-linear
least-squares regression algorithm (nlint, Matlab 2013a). Special attention was given to associate the rst mode of the distribution to the smaller
particles in the sample. This condition was imposed by enforcing the
mean value of rst mode to be smaller than that of the second (Eq. (3)).




1
1
b2 1
1 1
k1
k2

where is the Gamma function.

2.6. Experimental designs and statistical analysis
2.6.1. Role of gastric juice on the mechanical properties of the food matrix
The inuence of gastric juice on the mechanical properties of the potato matrix was investigated using a generalized randomized complete
block design.
To reduce the inuence of possible variations among the material
properties of the initial samples, the time of purchase of the potatoes
and the cooking batch were used as blocking factors. Potatoes bought
on the same day, and cooked in the same batch, were soaked in gastric
juice for up to 4 h, forming a block of experimental units as homogeneous as possible. The effect of gastric juice on the mechanical properties of the potatoes was investigated by measuring the material
properties of at least nine different potato cubes (subsamples) randomly selected after 0, 1, 2, and 4 h of soaking. The experimental design was
carried out in triplicates.
An analysis of variance (ANOVA) was performed using SAS 9.4 (SAS
Institute Inc., USA). A generalized randomized complete block model,
with random block effects was implemented (Mixed Procedure).
When needed, the mechanical properties of the potato matrix were
log-transformed to meet the assumptions of normality. A Tukey multiple comparison test was implemented to investigate the differences between all possible pairwise combinations of soaking times (LSMEANS
procedure, SAS).

K.C. Drechsler, M.J. Ferrua / Food Research International 88 (2016) 181190

185

Fig. 2. Characterization of broken particles into ne debris, fragments and undamaged product based on their relative size to the original one before comminution.

2.6.2. Role of gastric juice and compression levels on food breakdown

The inuence of soaking time and compression level on the breakdown behaviour of the potatoes was investigated using a split plot design.
The selection of soaking time as a main factor signicantly facilitated the
experimental analysis of the large number of samples involved.
Potatoes bought on the same day, and cooked in the same batch,
were soaked in gastric juice for a given and randomly selected period
of time. Soaked samples were then randomly exposed to each of the
four compression levels investigated, with three repeated observations
being performed at each level. The experiments were repeated until all
the levels of soaking times were investigated. The study was performed
in triplicates.
An analysis of variance (ANOVA) was performed on each of the parameters of the mixed Weibull distribution (SAS 9.4). A split plot design,
assuming random block effects was implemented (Mixed Procedure).
As before, dependent variables were log-transformed to meet the assumptions of normality when needed.

3. Results and discussion

3.1. Mechanical properties
As illustrated in Fig. 3, soaking in SGF had a major and detrimental
inuence on the mechanical properties of the potato matrix

(p b 1e 4). By the end of the soaking process, the failure strength

and toughness of the potato cubes decreased by over 80%, while their
apparent elasticity and surface hardness decreased by 66% and 40%,
respectively.
The major changes in the mechanical properties of the potato matrix
were observed within the rst 2 h of soaking, and were largely associated with the loss of intracellular cohesion and integrity of the tuber cell
walls by the acid hydrolysis of their pectin network (Moskowitz &
Kapsalis, 1976).
The relevance of these changes on the breakdown response of the
samples is discussed in the following sections.

3.2. Breakdown behaviour

3.2.1. Particle size distribution
As illustrated in Fig. 4, the distribution of particle sizes in the compressed potato samples varied signicantly depending on the parameter used to describe it (i.e., number of particles vs. particle surface area).
When described in terms of the number of particles, the distributions always exhibited a unimodal behaviour. Fitted to a Rosin
Rammler function (r2 N 0.98), these distributions suggested that the
samples were completely disintegrated into ne debris (A 7%A0), regardless of the soaking time and compression level analysed. On the
other hand, when described in terms of the particle surface area, the

Fig. 3. Changes in the mechanical properties of the potato matrix during soaking in simulated gastric uid (Estimate Standard error, SAS).

186

K.C. Drechsler, M.J. Ferrua / Food Research International 88 (2016) 181190

Fig. 4. Example illustrating the differences on the distribution of particle sizes depending on the parameter used to describe it: number of particles vs. surface area.

distributions always exhibited a bimodal behaviour, with one component associated with ne debris and another with particles similar in
size to the original cubes (N60% A0). Fitted to a mixed Weibull model
(r2 N 0.99), the bimodal behaviour of these distributions revealed the
signicant amount of product still present in the form of large particles
in the compressed samples. Even after 240 min of soaking in SGF and a
compression level of 80%, 35% (4% STD) of the compressed product
was still associated to particles similar in size to the original ones.
It is noteworthy the different information provided by the above distributions. Whether to use the number of particles or their surface area
to describe the breakdown behaviour of the samples will depend on
their relative ability to describe the physics of the process under study.
In oral processing, it is a common practice to describe the particle size
distribution of food boluses in terms of their number, as this information
is commonly associated with different mastication strategies and food
material properties. However, in the case of gastric digestion, it is not
particularly important to know how many particles of different sizes
are present in the digesta, as it is to know how much product (i.e., nutrients) is trapped within them. Unfortunately, when using 2D image analysis, the amount of product associated with different particle sizes
cannot be exactly determined. Nevertheless, the breakdown behaviour
of the samples was described by analyzing the size distribution of the
compressed particles in terms of their surface area (a better indicator
of the amount of product associated with them than their number).
In addition, the bimodal distribution of particle sizes during the trials
supports the idea that while the breakdown of foods at the end of an
oral/gastric process might be well described by a simple unimodal distribution, their breakdown behaviour during the process might not
(Mishellany et al., 2006, Kim et al., 2011).

3.2.2. Breakdown mechanisms

In the area of powder technology, the mechanisms underlying the
breakdown behaviour of a particulate system are determined based
on the relative amount of product associated with nes, fragments
and undamaged particles (Couroyer, Ghadiri, Laval, Brunard, &
Kolenda, 2000). However, unlike these classical applications, when
analysing the breakdown mechanisms of the potato samples, it was
found useful to consider not only the relative amount of product associated with different particle sizes, but also the number of particles associated with them.
Based on the distribution of product among different particle sizes,
the breakdown of the potatoes appears to be driven by surface damage
mechanisms (as only undamaged and ne debris particles are observed
within the samples, Fig. 5a). However, if surface effects controlled the
breakdown process, then the number of particles similar in size to the
original ones should be signicant; which as illustrated by Fig. 5b is
not the case. The only mechanism that can explain the information provided by both distributions is the collapsing of the potato matrix into
ne debris, without any progressive reduction in size due to fracture
or erosive mechanisms.

This result is further supported by an analysis into the size of the particles in which the potatoes broke down during the trials. Even without
any exposure to gastric juice and at compression level of only 20%, the
potatoes broke down into particles no larger than 28 m to 79 m.
Well within the size range of native starch granules (15100 m), the
size of these particles indicates a mechanical disintegration of the potato
matrix through the molecular disassembly of gelatinized starch granules formed during cooking.
Arguably obvious for the case under study, the above discussion
demonstrated a viable framework to investigate the underlying mechanisms of food breakdown.
3.2.3. Extent of particle damage and role of physicochemical effects
The breakdown behaviour of the compressed samples was characterized by tting the cumulative distributions of particle surface areas
to a mixed Weibull function (r2 N 0.99). The model parameters and correlation coefcients (r2) are given in Table 1.
In agreement with expectations, the level of soaking time and compressive load had a positive effect on the extent of damage incurred by
the particles (Fig. 6). In the following sections, the development of an
appropriate methodology to quantify the extent of damage incurred
by the potato samples is presented.
3.2.3.1. On the ability of A50 to describe particle damage. As mentioned in
Section 2.5.3, a parameter commonly used to characterize the breakdown behaviour of foods is the median value of their particle size distribution (A50).
The A50 of the compressed samples varied from 7.5 0.8 mm2 (1 h
soaking 20% compression) to 0.38 0.07 mm2 (4 h soaking 80%
compression). Values were signicantly inuenced by soaking time
(p = 5e 4), compression level (p b 1e 4), and an interaction between both factors (p = 5e 3). In general, the higher the soaking
time, and the higher the compression level, the smaller the value of
A50 (Fig. 7a).
While in agreement with expectations, it is noteworthy that changes
in A50 also suggested that for compression levels lower than 40%, neither soaking time nor compression level inuence the extent of particle
breakdown. Result that is in clear disagreement with the different proportion of ne particles observed in the samples under those conditions
(Fig. 7b).
As an alternative option, the extent of damage was characterized
based on the weight of ne particles within the compressed sample ().
As illustrated in Fig. 8, the relative weight of ne particles varied
from 0.09 0.05 (1 h soaking 20% compression) to 0.65 0.05
(4 h soaking 80% compression). Alike A50, this parameter was significantly affected by an interaction between soaking time and compression level (p = 0.01). However, unlike A50, changes in the values of
did reect the enhanced inuence of soaking time and compression
level on the extent of particle damage at compression levels lower
than 40% (Fig. 7b).

K.C. Drechsler, M.J. Ferrua / Food Research International 88 (2016) 181190

187

Fig. 5. Examples illustrating the information needed to determine the underlying mechanisms of food breakdown.

3.2.3.2. Role of chemical and mechanical effects on the breakdown process.

While a better parameter than A50 to characterize changes in the extent
of particle damage, the ability of to inform on the phenomena underlying these changes is still limited.
To gain a better insight into how different chemical and mechanical
conditions occurring during digestion might inuence the breakdown
process of the samples, changes in the distribution of ne and large particles within the samples were investigated.
An analysis of the scale and shape parameters of each Weibull mode
showed that while the distribution of small particles was signicantly
affected by soaking time and compression level (p b 0.006), the distribution of the large particles was not.
Exposure to gastric juice and increase compression levels not only
enhance the weight of ne particles within the samples (as discussed
above) but also shift their distribution in different manners (Fig. 9). As
compression levels increase, and more cubes collapse, the distribution
of nes shifted towards larger sizes until most of the cubes are damaged.
On the other hand, as exposure to gastric juice increases, the distribution of nes shifted towards smaller sizes due to the increased softening
of the potato matrix. A similar behaviour has been observed in the
breakdown of whey protein gels during in-vitro gastric digestion (Guo
et al., 2014), where the action of pepsin enhanced the breakdown of

the gels into smallest particles without any inuence on the size of the
larger particles present in the digesta.
In the case of large particles, while exposure to gastric juice and increased compression levels reduce their weight within the samples,
their size distribution remains largely unchanged (Fig. 9). This result
suggests that the breakdown process of large potato particles is largely
random and independent of the chemical and physical forces applied on
them. Result that is in good agreement with the breakdown mechanism
governing the process. As previously discussed in Section 3.2.2, the
potato cubes investigated in this study were already soft enough to disintegrate into ne debris under any of the compression levels analysed.
4. Conclusions
As part of a protocol aimed at investigating the underlying mechanisms of food disintegration during gastric digestion, this study developed
and tested a new methodology to inform on the role of gastric juice and
compressive forces on the breakdown mechanics of solid foods.
Using steamed potato cubes as a product model, this work revealed
the importance of the parameter used to describe the particle size distribution of the processed samples on the interpretation of their breakdown behaviour. It also suggested that food breakdown within the

Fig. 6. Graphical example of the inuence of soaking time and compression level on the cumulative size distribution of particle areas. Experimental averages (triplicates) STD.

188

K.C. Drechsler, M.J. Ferrua / Food Research International 88 (2016) 181190

Fig. 7. a) Median size of compressed particles, A50 (mm2). Estimated Standard error, SAS. b) Examples of the cumulative size distribution of particles areas after exposure to compression
levels lower than 40%. Experimental averages STD.

stomach might not follow a unimodal behaviour, highlighting the need

to characterize their behavior based on broad aspects of their particle
size distributions rather than the single point values commonly reported in literature. In this regard, this study demonstrated the applicability
of the mixed Weibull model to describe not only the breakdown behaviour of the samples, but also the relative role of the chemical and mechanical processes involved. A model commonly used to describe the
breakdown behaviour of a wide range of particulate systems across different disciplines, the general applicability of the mixed Weibull model
to describe the breakdown of solid foods during gastric digestion is
envisioned.
This study also demonstrated a viable methodology to determine
the contribution of erosive and fracture mechanisms to the breakdown
behaviour of the samples. We expect this methodology to be successfully applied to brittle products (such as almonds, apples, carrots) and soft,
deformable matrices as the one investigated in this study. However, its
applicability in the case of viscoelastic materials experiencing large deformations during the process (such as hydrogels or dry, amorphous
structures that can become ductile by moisture absorption or temperature effects) will require further analysis.
In addition, this study illustrates how an improved understanding of
the breakdown mechanics of solid foods under controlled conditions
can facilitate valid inferences with respect to their in-vivo performance
during digestion. In agreement with in-vitro data, this study shows
that while steamed potatoes can be easily disintegrated by the
compressing forces occurring in the stomach, their breakdown will
still be modulated by the combined action of the physicochemical

Fig. 8. Weight parameter (proportion of small particles within compressed samples).

Estimated Standard error, SAS.

processes involved. An analysis on the distribution of ne and large particles within the samples showed that while exposure to gastric juice
will likely not affect the breakdown probability of large particles, it
will indeed promote the disintegration of the food matrix into smaller
debris.
By analysing the relative role of chemical and physical processes on
the breakdown mechanics of solid foods during gastric digestion, the
proposed protocol would open new opportunities to identify the key
material properties underpinning the in-vivo performance of foods during digestion.
5. Current challenges and future considerations
A rst step in the development of a protocol to investigate the mechanisms of solid food disintegration during gastric digestion, this study
overlook a number of factors that need to be considered in future studies.
The inuence of oral processing is a critical step to be incorporated as
part of the proposed protocol. Changes in the particle size and material
properties of the ingested food during its mastication and saliva incorporation in the mouth would inuence its disintegration within the
stomach. Unfortunately, like in any other digestion process, the changes
occurring during oral processing are dependent on a number of food
and subject related factors difcult to characterize and model. By taking
into consideration that the proposed protocol is intended to investigate
the breakdown mechanics of solid foods during gastric digestion, rather
than its kinetics, we propose to simulate the inuence of oral processes
based on the standardized methodology proposed by Minekus et al.
(2014). The only change suggested to that methodology is to cut the
product into cubes of a given size, instead of mincing it before exposure
to the simulated saliva uid. As part of this study, 5 mm cubes were selected as a reasonable compromise between requirements for experimental analysis and the 0.1 to 6 mm size range observed in-vivo after
mastication (Hutchings et al., 2012; Jalabert-Malbos et al., 2007;
Mishellany et al., 2006; Peyron, Mishellany, & Woda, 2004).
Similar challenges will need to be addressed to standardize the invitro simulation of the chemical and physical processes occurring during
gastric digestion. Once in the stomach, foods are exposed to a wide and
dynamic range of physicochemical conditions, which again depend on a
number of food and subject related factors. The intrinsic variability of
these conditions, and the experimental difculty of their in-vivo analysis, makes the simulation of gastric digestion processes particularly difcult. In this study, the inuence of gastric juice on the chemical
degradation of the food matrix was investigated using a static in-vitro
digestion method, where samples were soaked in simulated gastric
juice at a constant pH of 1.8. While this methodology does not reect
the dynamic environment that food experienced in-vivo, it does provide
a standardized framework to investigate the mechanisms underlying
their chemical degradation during the process. For future studies, we
suggest following the methodology proposed by Minekus et al.

K.C. Drechsler, M.J. Ferrua / Food Research International 88 (2016) 181190

189

Fig. 9. Mean values of the size distribution of small and large particle within compressed samples (LSMean for each parameter under study Standard error, SAS).

(2014), where the pH value of the soaked samples is adjusted at a value

of 3.0. Nevertheless, the inuence of pH on the chemical degradation of
the food matrix could be a variable worth investigating, mostly if different pH conditions could inuence the molecular structure and physical
properties of the matrix. The simulation of the physical forces occurring
in the stomach is another area for future studies. While the bulk compression test used in this study led to forces similar in magnitude to
those recorded in-vivo, its ability to reproduce the shear stresses that
develop in-vivo is limited. Whether this is an important factor to
include, or whether the use of a simpler and widely available bulk compression test is enough to capture the mechanisms of particle breakdown during digestion, is an area for future studies.
Another important aspect to incorporate in this new protocol is an
analysis of the mechanisms driving the biochemical degradation of the
food matrix. In particular, those controlling the ability of gastric juice
to penetrate different food matrices and the consequent changes occurring on their structures. This analysis will be pivotal to identify how different material properties inuence the chemical degradation and
subsequent disintegration of foods during digestion; knowledge essential for the design of novel products with enhanced functionalities.
Finally, while not intended to predict the rate of food disintegration
during gastric digestion, the establishment of valid correlations between the outcomes of this protocol and the in-vivo behaviour of
foods during digestion will be essential. These correlations would not
only prove the value of the proposed protocol, but also help to classify
foods for their performance during digestion based on their relative response to the physicochemical processes involved.
Appendix A. Supplementary data
Supplementary data to this article can be found online at http://dx.
doi.org/10.1016/j.foodres.2016.02.019.
References
Abrahamsson, B., Albery, T., Eriksson, A., Gustafsson, I., & Sjberg, M. (2004). Food effects
on tablet disintegration. European Journal of Pharmaceutical Sciences, 22, 165172.

van Aken, G. A. (2010). Relating food emulsion structure and composition to the way it is
processed in the gastrointestinal tract and physiological responses: What are the opportunities? Food Biophysics, 5, 258283.
Bornhorst, G. M., Kostlan, K., & Singh, R. P. (2013a). Particle size distribution of brown and
white rice during gastric digestion measured by image analysis. Journal of Food
Science, 78(9), E1383E1391.
Bornhorst, G. M., & Singh, R. P. (2014). Gastric digestion in vivo and in vitro: How the
structural aspects of food inuence the digestion process. Annual Review of Food
Science and Technology, 5, 111132.
Bornhorst, G. M., Ferrua, M. J., Rutherfurd, S. M., Heldman, D. R., & Singh, R. P. (2013b).
Rheological properties and textural attributes of cooked brown and white rice during
gastric digestion in vivo. Food Biophysics, 8(2), 137150.
Brouwers, J., Tack, J., & Augustijns, P. (2007). Parallel monitoring of plasma and
intraluminal drug concentrations in man. After oral administration of fosamprenavir
in the fasted and fed state. Pharmaceutical Research, 24(10), 18621869.
Canet, W., Alvarez, M. D., & Gil, M. J. (2007). Fracture behaviour of potato samples (cv.
Desiree) under uniaxial compression. Journal of Food Engineering, 82, 427435.
Couroyer, C., Ghadiri, M., Laval, P., Brunard, N., & Kolenda, F. (2000). Methodology for investigating the mechanical strength of reforming catalyst beads. Oil & Gas Science and
Technology, 55(1), 6785.
Ghadiri, M., Ning, Z., Kenter, S. J., & Puik, E. (2000). Attrition of granular solids in a shear
cell. Chemical Engineering Science, 55, 54455456.
Guo, Q., Ye, A., Lad, M., Dalgleish, D., & Singh, H. (2014). Effect of gel structure on the gastric digestion of whey protein emulsion gels. Soft Matter, 10, 12141223.
Guo, Q., Ye, A., Lad, M., Dalgleish, D., & Singh, H. (2016). Impact of colloidal structure of
gastric digesta on in-vitro intestinal digestion of whey protein emulsion gels. Food
Hydrocolloids, 54, 255265.
Guo, Q., Ye, A., Lad, M., Ferrua, M., Dalgleish, D., & Singh, H. (2015). Disintegration kinetics
of food gels during gastric digestion and its role on gastric emptying: An in vitro analysis. Food and Function, 6(3), 756764.
Hutchings, S. C., Foster, K. D., Bronlund, J. E., Lentle, R. G., Jones, J. R., & Morgenstern, M. P.
(2012). Particle breakdown dynamics of heterogeneous foods during mastication: Peanuts embedded inside different food matrices. Journal of Food Engineering, 109, 736744.
Instron Engineering Corporation, Canton, MA (2015). Library. http://www.instron.com/
en-au/our-company/library/test-types/hardness-test/brinell-hardness-test?region=
Australia (Accessed Jun 2015)
Jalabert-Malbos, M. L., Mishellany-Dutour, A., Woda, A., & Peyron, M. A. (2007). Particle
size distribution in the food bolus after mastication of natural foods. Food Quality
and Preference, 18(5), 803812.
Kamba, M., Seta, Y., Kusai, A., Ikeda, M., & Nishimura, K. (2000). A unique dosage form to
evaluate the mechanical destructive force in gastrointestinal tract. International
Journal of Pharmaceutics, 208, 6170.
Kelly, K., O'Mahony, B., Lindsay, B., Jones, T., Grattan, T. J., Rostami-Hodjegan, A., ... Wilson,
C. G. (2003). Comparison of the rates of disintegration, gastric emptying, and drug absorption following administration of a new and a conventional paracetamol formulation, using gamma scintigraphy. Pharmaceutical Research, 20(10), 16681673.
Kim, E. H. J., Morgenstern, M. P., Bronlund, J. E., Fosterd, K. D., & Le Gote, A. (2011). Food
breakdown during human mastication Quantitative characterization. In P. S.

190

K.C. Drechsler, M.J. Ferrua / Food Research International 88 (2016) 181190

Taoukis, N. G. Stoforos, V. T. Karathanos, & G. D. Saravacos (Eds.), 11th International

Congress on Engineering and Food. vol. II, . Athens, Greece: Cosmosware.
Kong, F., & Singh, R. P. (2008). Disintegration of solid foods in human stomach. Journal of
Food Science, 73(5), R67R80.
Kong, F., & Singh, R. P. (2009). Modes of disintegration of solid foods in simulated gastric
environment. Food Biophysics, 4, 180190.
Koziolek, M., Grimm, M., Garbacz, G., Khn, J. P., & Weitschies, W. (2014). Intragastric volume changes after intake of a high-caloric, high-fat standard breakfast in healthy
human subjects investigated by MRI. Molecular Pharmaceutics, 11(5), 16321639.
Krifa, M. (2009). A mixed Weibull model for size reduction of particulate and brous materials. Powder Technology, 194, 233238.
Lamothe, S., Corbeil, M. M., Turgeon, S. L., & Britten, M. (2012). Inuence of cheese matrix
on lipid digestion in a simulated gastro-intestinal environment. Food and Function, 3,
724731.
Le Feunteun, S., Barb, F., Rmond, D., Mnard, O., Le Gouar, Y., Dupont, D., & Laroche, B.
(2014). Impact of the dairy matrix structure on milk protein digestion kinetics:
Mechanistic modelling based on mini-pig in vivo data. Food and Bioprocess
Technology, 7, 10991113.
Mackie, A. R., Raee, H., Malcolm, P., Salt, L., & van Aken, G. (2013). Specic food structures supress appetite through reduced gastric emptying rate. American Journal of
Physiology - Gastrointestinal and Liver Physiology, 304, G1038G1043.
Marciani, L., Gowland, P. A., Fillery-Travis, A., Manoj, P., Wright, J., Smith, A., ... Spiller, R. C.
(2001). Assessment of antral grinding of a model solid meal with echo-planar imaging. American Journal of Physiology - Gastrointestinal and Liver Physiology, 280(5),
G844G849.
Marciani, L., Hall, N., Pritchard, S. E., Cox, E. F., Totman, J. J., Lad, M., ... Spiller, R. C. (2012).
Preventing gastric sieving by blending a solid/water meal enhances satiation in
healthy humans. Journal of Nutrition, 142(7), 12531258.
Marciani, L., Pritchard, S. E., Hellier-Woods, C., Costigan, C., Hoad, C. L., Gowland, P. A., &
Spiller, R. C. (2013). Delayed gastric emptying and reduced postprandial small
bowel water content of equicaloric whole meal bread versus rice meals in healthy
subjects: Novel MRI insights. European Journal of Clinical Nutrition, 67, 754758.
Marciani, L., Wickham, M., Singh, G., Bush, D., Pick, B., Cox, E., ... Spiller, R. C. (2007). Enhancement of intragastric acid stability of a fat emulsion meal delays gastric emptying and increases cholecystokinin release and gallbladder contraction. American
Journal of Physiology - Gastrointestinal and Liver Physiology, 292(6), G1607G1613.

McClements, D. J., & Li, Y. (2010). Structured emulsion-based delivery systems: Controlling the digestion and release of lipophilic food components. Advances in Colloid
and Interface Science, 159, 213228.
Minekus, M., Alminger, M., Alvito, P., Ballance, S., Bohn, T., Bourlieu, C., ... Brodkorb, A.
(2014). A standardised static in vitro digestion method suitable for food An international consensus. Food and Function, 5, 11131124.
Mishellany, A., Woda, A., Labas, R., & Peyron, M. -A. (2006). The challenge of mastication:
preparing a bolus suitable for deglutition. Dysphagia, 8794.
Moskowitz, H. R., & Kapsalis, J. G. (1976). Psychophysical relations in texture. In J. M. De
Man, P. W. Voisey, V. F. Rasper, & D. W. Stanley (Eds.), Rheology and texture in food
quality (pp. 554). Connecticut: AVI Publishing Co.
Mllertz, A. (2010). Food effects on drug absorption and dosage form performance. In J. B.
Dressman, & C. Reppas (Eds.), Oral drug absorption. Prediction and assessment
(pp. 101). Florida: Taylor & Francis Group.
Nicolai, T., & Durand, D. (2013). Controlled food protein aggregation for new functionality.
Current Opinion in Colloid & Interface Science, 18(4), 249256.
Norton, J. E., Wallis, G. A., Spyropoulos, F., Lillford, P. J., & Norton, I. T. (2014). Designing
food structures for nutritional and health benets. Annual Reviews of Food Science
and Technology, 5, 177195.
Parada, J., & Aguilera, J. M. (2007). Food microstructure affects the bioavailability of several nutrients. Journal of Food Science, 72(2), R21R32.
Peyron, M. -A., Mishellany, A., & Woda, A. (2004). Particle size distribution of food boluses
after mastication of six natural foods. Journal of Dental Research, 83(7), 578582.
Turgeon, S. L., & Rioux, L. E. (2011). Food matrix impact on macronutrients nutritional
properties. Food Hydrocolloids, 25(8), 19151924.
Urbain, J. L., Siegel, J. A., Charkes, N. D., Maurer, A. H., Malmud, L. S., & Fisher, R. S. (1989).
The two-component stomach: Effects of meal particle size on fundal and antral emptying. European Journal of Nuclear Medicine, 15(5), 254259.
Van Kleef, E., Van Trijp, J. C. M., Van Den Borne, J. J. G. C., & Zondervan, C. (2012). Successful development of satiety enhancing food products: Towards a multidisciplinary
agenda of research challenges. Critical Reviews in Food Science and Nutrition, 52,
611628.
Walter-Sack, I. E., de Vries, J. X., Nickel, B., Stenzhorn, G., & Weber, E. (1989). The inuence
of different formula diets and different pharmaceutical formulations on the systemic
availability of paracetamol, gallbladder size, and plasma glucose. International Journal
of Clinical Pharmacology, Therapy and Toxicology, 27(11), 544550.