Luuk P. A. Simons
and J. Felix Hampe
Computer Science, Delft University of Technology, Mekelweg 4, Delft, Netherlands
Institute for IS research, Koblenz University, Koblenz, Germany
Keywords: Smart phone, Mobile application, Lifestyle intervention, Health, Motivation, Patient Support, Prostate
Cancer, Service design, Design research.
Abstract: There is an existing Health (e)Coach Solution for supporting intensive lifestyle changes, which may help
reduce cancer progression risks in men with low grade prostate cancer. An important challenge is to support
and motivate healthy behaviors for the long term (4 years). Smart phones are increasingly used also in this
age group. And mobile apps offer significant opportunities for personalized health behavior monitoring and
support. Our overall proposition is that an mApps suite, as an extension to an existing online dashboard with
automated emails and interpersonal coach sessions, appears promising for improving long term health
behavior support. That is, if three conditions are met: 1) Using best of breed mApps from the market
(benefits like ease-of-use, continuity, customer support); 2) creating easy and meaningful score conversions
(from mApps to personal dashboard); 3) using automated emails as glue between three key aspects personal
progress/dashboarding, user attention and motivation, mApp usage.As methodological approach, this paper
focuses on two phases from a design research cycle: requirements analysis and a preliminary design solution
exploration. In the analysis phase, two questions are addressed: What are effective lifestyle intervention
components according to literature? What are solution properties that may enhance motivation and improve
health behaviors? In the solution exploration phase our design question is how a proposed mApp extension
may likely add value to the existing Health (e)Coach solution.
This paper bridges the two worlds of health and ICT
(Information & Communication Technology). After
the methodology introduction, the first part of the
paper explores the extent of potential benefits from
extensive lifestyle interventions for prostate cancer
patients. The second part analyzes requirements for
a proposed extension with smart phone apps in
relation to an existing health (e)Coach solution. This
is partly based on literature and partly on interviews
with the respective health solution provider. Finally,
we describe a preliminary solution.
1.1 Propositions
In summary, our overall proposition is that a smart
phone applications (mApps) suite, as an extension to
an existing online dashboard with automated emails
(i.e. push notification) and interpersonal coach
sessions, appears promising for improving long term
health behavior support for low grade prostate
cancer patients.
This builds on four sub-propositions: 1) A gain
may be achieved in quantity and quality of life for
low grade prostate cancer by enhancing existing
treatments with an intensive healthy lifestyle
intervention. 2) An existing health (e)coach
program, which uses personal e-dashboarding,
intensive health education, interpersonal coaching
and automated personal progress emails, offers
promising results in terms of health beliefs, health
behaviors and health results (Simons and Hampe
2011). Still, more value could be added by bringing
the health experiences and feedback even closer to
the users. 3) Adding a portfolio of best of breed
existing smart phone apps may help improve health
awareness, quality of health beliefs, help improve
health behaviors, increase transparency, fun and
motivation, as well as long term healthy lifestyle
adherence. 4) This rests, however, on a number of
conditions: a) Using best of breed existing smart
phone applications from the market. This brings
P. A. Simons L. and Felix Hampe J..
DOI: 10.5220/0003872704430448
In Proceedings of the International Conference on Health Informatics (HEALTHINF-2012), pages 443-448
ISBN: 978-989-8425-88-1
2012 SCITEPRESS (Science and Technology Publications, Lda.)
benefits like attractiveness, ease-of-use, low cost
implementation, flexible adaptation, continuity, and
high quality customer support per application. b)
Creating easy and meaningful conversions from the
mApps sub-scores to the health progress scores in
the personal dashboard. For example: How many
calories have been burned with intensive exercise
this week? Or: How may fibers or fats have been
consumed this week? The mobile applications in the
portfolio must be selected to provide these types of
sub-scores which can then be translated into the
overall health behavior progress scores in the
personal dashboard. c) Using automated emails as
glue between three key aspects of the solution:
facilitating personal progress feedback (based on the
e-dashboarding), increasing user health attention and
motivation, and automated feedback or compliments
on mApp usage. In other words: using the mApps
must not be a stand-alone activity. It should visibly
increase the value of the personal dashboard and
increase health behavior motivation. Whether
patients already are active mApp users or not, the
weekly email notifications they receive should
integrate and explain their personal e-dashboard
scores (and when active: mApp scores).
Furthermore, this channel mixing increases
awareness and it may motivate non-users to start
using mApps by showing the benefits this generates.
1.2 Design Research Cycle
Our research approach is based on Information
Systems Engineering. It uses the design research
cycle of Vaishnavi and Kuechler (2004) with the
enhancements towards a pluralistic view as
discussed in Frank (2006). Thus we describe the first
two stages of the design cycle (Vaishnavi and
Kuechler, 2004) in this paper. That is: problem
analysis and a motivated suggestion for a design
solution for the problem. For the service concept
design and service implementation only initial
evaluation results are available, a large scale
evaluation is missing. Thus we present research in
progress. Nonetheless, we propose the general
positions and the examplification along the prostate
cancer service platform for discussion and
improvement to the research community.
2.1 Large Differences Worldwide
There are large incidence differences worldwide for
several of our common cancers: lung cancer,
colorectal cancer, children’s cancers, as well as
breast cancer and prostate cancer. For prostate
cancer, there is a 100-fold difference in age-
corrected incidence rates. These vary from
0,5/100.000 persons per year in Qidong County in
China in 1997 to around 125/100.000 persons per
year in the USA.
These differences are strongly correlated with an
affluent (Western) lifestyle and diet. Degree of
physical activity, stress and environmental factors
likely play a role, but the hypothesis that diet has an
independent effect is supported by migration studies
which show that Asians who maintain an Asian food
pattern through adolescence while living in the West
have better cancer protection than those who have
adopted Western foods (Wu, 2002).
2.2 Lifestyle Factors
The World Health Organization defines health as
consisting of physical, mental and social health. And
each of these components has an effect on prostate
cancer incidence and mortality. Below we review
studies on four aspects: social support, stress,
physical activity and food.
2.2.1 Social Support
Furthest away from the field of well-grounded
biological pathways is the empirical data regarding
social support and cancers. (Various immune-,
hormonal-, neural- and also behavioral mechanisms
have been proposed, but this field is still young.) On
the one hand there have been several long term
prospective studies over the past decades which have
shown correlations between degree of social support
early in life and cancer incidence and mortality years
later. For example the Alameda study measured
degree of social support among Alameda County
residents and found significantly raised cancer risks
and mortality of all sites during the 17-year follow
up period (Berkman, 1995); (Reynolds and Kaplan,
1990). They found it to be an independent and more
powerful predictor than age, gender, socioeconomic
status and health behaviors. Also marital status has
an impact on five year survival for almost every
category of age, gender and stage of various cancers
(Goodwin, 1987). This was also found specifically
for prostate cancer (Krongrad, 1996). These effects
have been confirmed in recent meta-analyses:
Across 87 controlled studies, having high levels of
perceived social support, larger social network, and
being married were associated with decreases in
HEALTHINF 2012 - International Conference on Health Informatics
relative risk for mortality of 25%, 20%, and 12%,
respectively. There were stronger associations for
leukemia/lymphomas and breast cancer, but not
enough studies for other cancers to differentiate
effect sizes (Pinquart and Duberstein, 2010). And as
a generic health effect: across 148 studies, the effect
size was OR = 1.50 (95% CI 1.42 to 1.59), 50%
more likelihood of survival with stronger social
relationships, a finding consistent across age, sex,
initial health status, cause of death, and follow-up
period (Holt-Lunstad, 2010).
On the other hand, there is the question how
effective psycho-social interventions are. There was
initial enthusiasm based on successful interventions
(Spiegel, 1989); (Fawzy, 1993), but larger RCTs
have not been able to replicate these results
(Kissane, 2007). Still, it is interesting that in one
study, both control and treatment groups had half the
mortality risk of those who declined to participate in
the study (Boesen, 2007). And another study showed
that those who reduced their degree of depression in
the first year had improved survival, irrespective of
being in the control or treatment group (Giese-
Davis, 2011). An important complexity in this field
is that seeking support has become very normal, and
cancer patients may do so independent of a formal
support group. For example, even the simple yes/no
answer to the question ‘Do you have someone to talk
to in the 3 months after your diagnosis?’ correlated
with a 30% increase (for ‘yes’) in survival after 7
years (Maunsell, 1995). And a review of the psycho-
oncological literature concluded (Garssen, 2004)
that one of the most consistent findings is that: the
more hopelessness and helplessness are experienced,
the more unfavorable the cancer progression is
(Watson 1999); (Goldberg, 1996); (Schulz 1996).
2.2.2 Stress
A next potential factor is the effect of chronic
stress. Potential pathways run via increased cell
aging and and telomere shortening (Epel, 2004);
(Puterman, 2010), increased DNA damage and
hampered DNA repair (Flint, 2007), reduced
immune function (Segerstrom, 2004), increased
levels of inflammatory cytokines (Bower, 2007), and
stress-induced activation of the sympathetic nervous
system (SNS) and hypothalamic–pituitary–adrenal
(HPA) axis which results in the production of
catecholamines and cortisol, which have direct
effects on epithelial cell growth and tumor
vascularization (McGregor, 2009). Besides, there are
higher order mechanisms like stress-induced
depression (for example, depression raised mortality
in men for melanoma, colon and prostate cancer
with 67%: Almeida 2010), reduced self-efficacy and
self-care (Miller, 2001) or the 90% increased cancer
risk due to low perceived quality of life (Flensborg-
Madsen, 2011). For an overview of biological
mechanisms, see McGregor (2009). Next, one study
found that men who experienced high levels of
stress were more than three times as likely to have
elevated PSA levels (Prostate Specific Antigen) than
were men who experienced low levels of stress
(Stone, 1999). And using the breast/prostate cancer
analogy, another large study showed that African
American women who experienced more emotional
stress due to discrimination (on the job, on the street
and by the police) had a 48% increased risk of
developing cancer after six years (Taylor, 2007).
In terms of interventions, cognitive-behavioral
therapy (Philips, 2008) and well as yoga
(Raghavendra, 2009) and meditation programs
(Biegler, 2009) have resulted in lower cortisol
levels, as well as improved mental health and quality
of life (Nidich, 2009); (Beard, 2011); (Chambers,
2011). Survival impacts are not always clear or
made explicit. When they do appear they seem
dependent on improvements in relation to depression
(Giese-Davis, 2011) hopelessness (Garssen, 2004) or
quality of life (Flensborg-Madsen, 2011); (Spiegel
2.2.3 Physical Activity
Two different effects exist between physical activity
and cancer. The first effect is that sedentary behavior
increases cancer risk, independent of BMI (Body
Mass Index) or other activity. Sedentary behavior is
associated with adverse cardio-metabolic and cancer
mortality; in a review of 18 cancer studies has been
associated with increased colorectal, endometrial,
ovarian, and prostate cancer risk (Lynch, 2010). The
second effect is that vigorous exercise helps reduce
cancer risk. Already in the College Study, running
for over 40 years and following 50.000+
participants, a more than 30% reduction in cancer
mortality risk due to physical exercise across all
cancers was found, including prostate cancer when
exercise is heavy (Paffenbarger, 1986). As one
hypothesized pathway, heavy exercise decreases
testosterone, which in turn is linked to lower prostate
cancer risk.
The EPIC study found an inverse association
between advanced prostate cancer risk and
occupational physical activity - which is inversely
correlated with inactivity throughout the day - but
not with leisure activity (Johnsen, 2009). A study
specifically designed for investigating the impact on
survival of higher physical activity after prostate
cancer diagnosis, found 46% lower overall mortality
with increased activity and a 61% lower prostate
cancer mortality for those men who had 3 or more
hours weekly of vigorous exercise compared with
less than 1 hour weekly (Kenfield, 2011). Men
exercising vigorously before and after diagnosis had
the lowest risk. And a study focusing on degree of
progression found that men who walked briskly for
3 h/wk or more had a 57% lower rate of progression
than men who walked at an easy pace for less than 3
h/wk (Richman, 2011). Other advantages from
physical activity for prostate cancer survivors are
improved muscular fitness, physical functioning,
fatigue, and quality of life (Thorsen, 2008).
2.2.4 Dietary Factors
One of the best known factors between lifestyle, diet
and cancer is obesity. This link has been reviewed
extensively in the WCRF Research report (WCRF,
2007). Hypothesized pathways include: pro-
inflammatory and angiogenic effects of adipose
tissue, increased oxidative stress on cells in obese
people, and increased blood serum levels in obese
people of insulin, glucose and lipids, which all tend
to promote cancer cell growth. For various cancers,
increased serum levels of insulin, glucose and lipids
as caused by dietary fat, have been shown to
increase tumor growth and aggressiveness,
independent of BMI. For prostate cancer the
evidence is still sparse. Glycemic load appears to
have limited impact (Nimptsch, 2011), but some
protective effects from low fat diets have been found
(Aronson, 2010); (Kobayashi, 2008).
The obesity evidence is ‘convincing’ (WCRF
2007) for cancers of the oesophagus, pancreas,
colorectum, breast (postmenopause), endometrium
and kidney. The link between obesity and prostate
cancer is not free of controversy however. Several
recent studies suggest that obesity increases the risks
of prostate cancer aggressiveness and decreases the
risks in relation to low grade cancers (Gong, 2006);
(Freedland, 2008). Freedland et al go on to suggest
that obesity is biologically associated with increased
risk, although this is obscured owing to
hemodilution of prostate-specific antigen (PSA) and
larger prostate size (which causes cancer-
underestimations during diagnosis). This biological
link was confirmed in another large study showing
that men who lost weight reduced their prostate
cancer risk and those with high BMI were at
increased risk for aggressive prostate cancer
(Rodriguez, 2007).
Chronic inflammation is also relevant: the
(modified) Glasgow Prognosic Score, built on CRP
(C-Reactive Protein) and albumin measurements,
but also other inflammation-based prognostic scores,
predict cancer survival, independent of age, sex,
tumor stage and for all tumor sites. This includes
common cancers like colorectal, breast, lung and
prostate cancer (Proctor 2011-1, Proctor 2011-2).
This may be a protective link from which
fruits/vegetables (Cohen, 2000); (Hodge 2004) and
flaxseed/omega 3 PUFA here, besides reducing
testosterone and androgens (Demark-Wahnefried,
2001; 2008). Recently it has been found that also
aggressive prostate cancer may be attenuated with
leafy (-34%) and high carotenoid vegetables (-29%)
and promoted (+64%) with high glycemic load foods
(Hardin, 2011). Inflammation and CRP levels can be
reduced via diet (Middleton 1998), as well as via
physical activity (Ford, 2002); (Wegge, 2004) and
stress reduction (Fang, 2010); (Oh, 2010).
Furthermore, healthy diet can also help reduce
depression (Akbaraly, 2009); (Jacka, 2010);
(Beezhold, 2010).
Most people are not aware that increased blood
cholesterol levels also indicate an increased cancer
risk, besides increased heart disease risk. This has
been shown for several cancers. For prostate cancer
there are indications that lower cholesterol levels
reduces incidence of the more advanced stages of
prostate cancer (Platz, 2009). Unfortunately,
pathways are still unclear. Still, diet can create large
improvements in cholesterol levels, as can physical
activity and stress management.
Western diets (rich in animal foods and low in
fiber) also tend to raise testosterone levels (Ross,
1994); (Habito, 2001) just like they raise estrogen
levels in women (Campbell, 2006). And increased
serum testosterone is a known prostate cancer risk
A fifth dietary factor is IGF-1, in concert with a
wider growth factor family which promote cancer
growth. For a review of 194 publications, see Pollak
2008. High circulating levels of IGF’s and low
levels of its binding protein IGFPB-3 are associated
with increased risk of several cancers, including
those of the breast, colorectum (Ma, 2001), lung
(Yu, 1999) and prostate (Chan, 2002); (Chan, 1998).
IGF-1 levels are relatively high in the West, and our
meats and dairy products have seen their IGF-1
levels double over the past decades, whereas IGF-1
has been shown to reach the receptor sites in the gut
in its biologically active form (EU Scientific
HEALTHINF 2012 - International Conference on Health Informatics
Committee, 1999). Blood levels of IGF-1 are
influenced via diet in multiple ways: most notably
via milk, as shown in cohort studies in men (WCRF,
2007); Ma, 2009) and women (Norat, 2007) as well
as experimentally: for example a Danish study found
a 30% increase in circulating IGF-1 levels when
increasing daily milk consumption from 200 ml to
600 ml daily. Besides IGF-1 increases from dairy
and more mildly from other animal proteins, IGF-1
reductions were found to correlate with increased
vegetable intake (Gunnell, 2003); Norat, 2007).
As a further connection between diet and cancer,
there are the anti-angiogenic properties of foods.
The research groups of Folkman and Li have
categorized many food items according to anti-
angiogenic strength. It turns out that many foods
high in micronutrients (like berries, vegetables,
fruits, spices) also potently block the growth of
blood vessels towards microtumors. See food items
and effect sizes being compared to anti-cancer drugs Hence, the
microtumors will not become full blown tumors.
The anti-angiogenic effect sizes appear similar to
several of the recent FDA approved anti-angiogenic
drugs, based on the zebrafish vessel formation tests
(He, 2009).
On the level of specific dietary product groups
and their effects on prostate cancer, the WCRF
(2007, p 306-308) reports the following:
On the basis of 20 studies on pulses (legumes)
and 10 studies on soy and soy products: ‘Most
studies showed decreased risk with increased intake.
Meta-analysis of case-control data produced
evidence of an association with legume intake, with
a clear dose-response relationship.’
Ten studies investigated processed meat: ’All
cohort studies reported increased risk with higher
intake; and most case-control studies also showed
this effect.’
25 Studies investigated milk and dairy and 38
studies investigated milk: ‘Most of the studies
showed increased risk with increased intake. Meta-
analysis of cohort data produced evidence of a clear
dose-response relationship between advanced/
aggressive cancer risk with milk intake and between
all prostate cancer risk and milk and dairy intake.’
The WCRF report mentions 2 pathways between
milk and prostate cancer: via the IGF axis, and via
downregulation of bioactive vitamin D, thereby
increasing cell proliferation in precancerous cells.
23 Studies investigated dietary calcium (marker
for dairy intake): ‘Most [of the 9] cohort studies
showed in creased risk with increased calcium
intake; case control studies were inconsistent. Meta-
analysis of cohort data showed an increased risk of
27% per g/day; [and] an increased risk of 32% per
g/day [for advanced or aggressive prostate cancer]’
On the basis of 10 studies measuring dietary
selenium and 23 studies measuring in vivo selenium
levels: ‘Most studies, including those reporting
separately on advanced/aggressive prostate cancer,
showed decreased risk with increased intake.’
On the basis of 42 studies on foods containing
lycopene: ‘Most of the studies showed decreased
risk with increased intake. Meta-analysis [ .. ]
showed a 4% decreased risk per 10 microgram
lycopene/liter serum.’
Foods containing vitamin E, in 38 studies: ‘Most
studies showed decreased risk with increased
2.2.5 Summary
In summary, social support, stress, physical activity
and foods each may have positive influences on
mental and physical health. A clinical study with
low grade prostate cancer patients is in preparation
where these factors will be targets for a lifestyle
Elsewhere (Simons and Hampe, 2011) an existing
health (e)coach solution is described which supports
intensive lifestyle change addressing each of the four
lifestyle components: social support, stress, physical
activity and food. This solution strongly builds on
empowering patients to improve their health:
“Health is what happens between doctor’s visits.”
The clinicians and the solution provider are
interested in adding an mApp portfolio to improve
closeness and feedback to the patients.
The purpose of this section is to formulate a
preliminary requirements and solution properties
list. Requirements are extracted from literature as
well as from two interviews with the solution
provider (health coach and innovation manager).
The latter interviews also led to the formulation of
the preliminary mApp based solution concept.
3.1 Requirements
According to the meta-study of Jimison et al., (2008)
a number of requirements must be fulfilled to create
ICT solutions that enhance the value for patients: 1)
Ease of use, 2) Perceived usefulness, 3) Relevant
information loop: perceived as having health
relevance by the patients, 4) Embedding in a
treatment cycle: interaction with health providers to
decide on the course of action following the
information which has been generated, 5) Delivered
on technology that patients use every day.
Based on the health providers’ interviews, a
central goal in this intensive lifestyle intervention is
health empowerment and involvement. Related
requirements are: 6) Improving transparency and
awareness of health effects from lifestyle behaviors,
7) Improving quality of health beliefs, 8) Increasing
fun and attractiveness of tracking health behaviors,
9) Increasing motivation for health behaviors.
From the supply side of the health (e)coach
provider, the following requirements were
mentioned: 10) Operational quality and reliability
(customer service of mApps, reliability, continuity,
ease of use), 11) Low cost implementation and
flexibility (simple to test, adopt and discard mApps,
thus gradually improving the portfolio), 12) Creating
meaningful conversion scores from mApps into
existing e-dashboard, 13) Information flows
integration (from mApp to e-dashboard to coach and
vice versa).
3.2 Preliminary Solution Concept
In relation to user needs, key elements of the
prospected solution are ‘closeness’, attractiveness
and relevance: everyday interfaces are to be used,
like email and smart phones. Also fun and
attractiveness are important, hence the provider
choice for selecting mApps from the most popular,
existing mApps. In Aug 2011, smart phone
penetration in the Netherlands was 42% http://on
In the age group of 60-70 years (core of this patient
group) penetration was 20+% and rapidly rising.
Also for this age group, smart phones may become a
dominant personal ICT platform (in close
competition with email, internet and telephone),
because of their attractiveness, 24x7 closeness, and
intuitive user interfaces.
In terms of functionality, for each health factor
(food, physical activity, stress, social support) one or
several mApps are selected. Table 1 gives an
overview of the functionality and design rationale.
The selected mApps are offered and supported as
a portfolio, with a list of recommended mApps, a
‘runner up of the month’ that is recommended to try
out, with a Dutch user manual, and with meta-data
generated weekly for the patients (delivered in an
automated email) to stimulate mApps usage and
show mApps benefits which have been experienced
by several of the users, to the rest of the patient
group. When patients open their e-dashboard, they
see references to the mApps portfolio. And when
they enter their weekly health behaviors in the e-
dashboard, they see the options to enter mApp
generated values.
Table 1: Preliminary Functionality and Rationale.
Functionality Design rationale
Everyday smartphone & mailbox use.
Increase ease of use and
Portfolio of mApp support options:
stress, exercise, food, group support.
Something useful for
Selected mApps are managed as
portfolio, including user manual.
Manage and monitor mApps
use. And explain value
adding in relation to e-
dashboarding and coaching
Explicit links into dashboard Æ if
XYZ calories burned: input that
value; input from week food report:
fibers etc.
Visualize cross-referencing
and mutual roles of e-
dashboard versus mApps
Explicit references in dashboard to
mApps: pictures, explanations and
dropdowns: ‘input value from App’.
Visualize options for mApps
use, and trigger non-users to
consider using mApps.
Using weekly low intrusion mails
(including reports from mHealth data
Give progress feedback, add
relevance, increase
Based on the preceding argument, our conclusion is
as follows: An mApps suite, as an extension to an
existing online dashboard with automated emails
and interpersonal coach sessions, appears promising
for improving long term health behavior support.
This rests on the sub-propositions summarized on
page 1. Furthermore, we expect that in the coming
ten years mApps will revolutionize health self-
management, as well as transform the roles and
activities of health providers. They will increasingly
become coaches of (information- and mApp-)
empowered patients that “make health happen in
between doctor’s visits”.
--Truncated – Reference list available with authors --
HEALTHINF 2012 - International Conference on Health Informatics