Hindawi Publishing Corporation

Journal of Obesity
Volume 2012, Article ID 546459, 11 pages
doi:10.1155/2012/546459

Review Article
Physical Activity Behaviour: An Overview of Current and
Emergent Theoretical Practices

Duncan S. Buchan,1 Stewart Ollis,1 Non E. Thomas,2 and Julien S. Baker1

1 Health and Exercise Sciences, School of Science and Technology, University of the West of Scotland, Hamilton ML3 0JB, UK
2 School of Human Sciences, Swansea University, Swansea SA2 8PP, UK

Correspondence should be addressed to Duncan S. Buchan, [email protected]

Received 21 November 2011; Revised 30 March 2012; Accepted 10 May 2012

Academic Editor: Amy A. Gorin

Copyright © 2012 Duncan S. Buchan et al. This is an open access article distributed under the Creative Commons Attribution
License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly
cited.

Physical activity research has been dominated by traditional cognitive rationale paradigms utilized within other domains. Though
this approach to physical activity behavior has greatly enhanced our understanding of the key determinants, it has done little to
eradicate the health problems we currently face. In order to achieve lasting change though, multilevel interventions may prove
effective. Ecological perspectives have been proposed as an effective approach in combating current physical inactivity levels.
Nevertheless, this approach is in its infancy and much has still to be learned. The aim of this paper is to provide an overview
of the main behavioral models used within the physical activity domain while proposing the need for further models that will
embrace the principles presented by ecological and complexity theories.

1. Introduction Theory of Planned Behavior (TPB), The Self-Determination

Theory (SDT), and The Transtheoretical Model (TTM) [5].
The World Health Organization (WHO) rates physical Undoubtedly the application of these theories has greatly
inactivity as one of the main causes of premature death in enhanced our understanding of the psychological influences
developed countries, implicated in the aetiology of many and processes which influence PA behaviour. Even so, this
chronic diseases such as cancer, cardiovascular disease, reliance upon traditional cognitive rationale paradigms has
diabetes, and obesity [1]. Though the beneficial effects of done little to eradicate current health problems.
physical activity (PA) on health are well known and firmly What has become apparent within the last decade is
established, few are meeting current PA recommendations that changing behavior is a complex and multifaceted phe-
[2]. It is now appreciated that individuals face considerable nomenon with multiple levels of influences. Thus, in
barriers when changing complex behaviors such as PA [3]. order to achieve enduring changes in behavior multilevel
Though early endeavours into the promotion of PA tended interventions that focus on targeting individuals, social
to be largely atheoretical, the need for interventions to be environments, physical environments, and policies have
informed through appropriate theoretical underpinning and been proposed [6]. Since social ecological models of health
allow subsequent replication was essential. This paradigm behaviour focus on individual, as well as social, policy, and
shift saw subsequent authors focusing towards understand- environmental influences [7, 8], researchers have recently
ing the determinants and correlates of PA, in particular embraced the use of such frameworks to inform interven-
psychosocial influences [4]. As such, theories of behavioural tions. It seems that, once again, the PA domain is at the
change initially developed within social psychology have tipping point of another paradigm shift.
dominated the literature. Despite ecologically based multilevel interventions hold-
Though some models have been applied more frequently ing great potential in changing complex behaviors, cer-
than others, the four most prominent theories utilized within tain issues associated with such an approach needs to be
a PA context are The Social Cognitive Theory (SCT), The considered. For instance, though guiding frameworks such
2 Journal of Obesity

as the social ecological theory are useful for considering discuss the self-efficacy theory and consider its predictive
behavioural determinants broadly, there is a lack of models utility and application within the PA domain.
available that provide specific mechanisms through which
particular influences may interact and influence behavior.
This is to be expected given that much research is still needed 4. The Self-Efficacy Theory within
before an agreed consensus regarding the factors that exert Physical Activity
the most influence on PA behavior, in a specific context, is
formed. As the TTM and the TPB provide mechanisms and Since its conception over 30 years ago literally hundreds of
variables to target for influencing behavior, it is unsurprising PA-related research has utilised the construct of self-efficacy
that researchers have tended to rely heavily upon these as an antecedent, outcome, or process variable when trying
theories when devising interventions. Thus, the purpose of to understand the motivational process involved in either
this paper is twofold. First, a very brief critique of four sporting or exercise performance [15]. The reason for its
popular cognitive-based theories used within the PA domain popularity is twofold. As discussed, self-efficacy for PA has
will be presented. Secondly, attention will be given to the been shown to predict those individuals who will engage
emergent ecological models currently being proposed with in such endeavours [14] but of greater value is that the
further research directions being offered. construct of self-efficacy comes with specific guidelines for its
development. This undoubtedly has appealing qualities for
practitioners and researchers which may explain popularity
2. Current Theoretical Practice within of this construct in understanding PA behavior.
Physical Activity There is a diverse body of literature that has utilised
the self-efficacy theory within a variety of health and PA
The two distinct approaches currently dominating the
contexts, such as weight loss [16, 17], exercise in older
literature are the staged-based approach and the cognitive-
adults [18], and exercise in adolescent girls [14]. Specifically,
based approach [4]. Whereas stage-based models propose
self-efficacy is thought to influence the goals people set,
that individuals go through stages in order to adopt or main-
their ability to persist in the face of obstacles, and their
tain complex behaviors like PA, cognitive-based approaches
capacity to cope with setbacks and stress and as such, directly
assume that complex behaviors are controlled by rationale
influence behavioral engagement. Evidence is now clear in
cognitive activity. Within this approach interventions focus
support of this. For instance, Dishman et al. [14] evaluated
on indentifying determinants that can explain behavior. At
the effects of a school-based intervention on variables of
this point it is important to recognize that other popular
the SCT, designed to emphasise changes on instruction
models exist including the Sports Commitment Model [9],
and the school environment. The authors found that the
the Schema Theory, the Psychological Continuum Model,
manipulation of self-efficacy had a direct increase in PA
the Social Support Model, the Enjoyment Model, and the
levels among adolescent girls and encouraged the use of self-
Health Belief Model [6]. However, the TPB, the SCT, the
efficacy as a variable that should be targeted to raise PA
SDT, and The TTM models represent those which have been
levels within this population group. Similarly, self-efficacy for
tested and adopted most widely in the health behavior and
PA has been shown to predict engagement in walking and
PA literature [10]. Therefore it is felt that these four models
identify adherers and dropouts in PA interventions [14, 16].
are most worthy of attention and will be discussed within this
While the evidence clearly supports the use of self-
brief paper.
efficacy as a powerful predictor of behavior, further research
has been suggested, particularly in regards to the mea-
3. The Social Cognitive Theory within surement of self-efficacy. As proposed within the theory,
Physical Activity self-efficacy will vary along the dimensions of magnitude,
strength, and generality. Though previous research has
A widely used theoretical model of behavior change favoured measured the strength of self-efficacy [12, 14], few have
by researchers is the SCT [11]. This model describes factors included measures of magnitude and generality. As such,
that may affect and determine behavior while also specifying the predictive utility of self-efficacy may be misinterpreted.
mechanisms through which these factors work and how For instance, consider the findings of Dishman et al. [14]
they may be altered into effective health behaviors. The who noted that increased self-efficacy, albeit small, had a
structure and predictive utility of the SCT has been tested direct effect on increasing PA behavior. The authors included
across numerous domains and populations [12, 13] and has a variety of activities deemed popular with high school
emerged as one of the most prominent frameworks adopted girls such as aerobics, weight training, dance, and self-
in the study of motivation and behavioral outcomes. Con- defence classes but failed to measure the generality of efficacy
sistently, research has shown self-efficacy as a key variable between each of these situations.
within the SCT and is said to be the most powerful factor It could be that the level of efficacy for participants
to consider when predicting behavior [12]. Surprisingly, was high when performing aerobics but relatively small
little research has involved the application of the SCT in when performing weight training. Providing an overall
its entirety within the PA domain. Of the research that has, measurement of efficacy from the four activities, and not
only self-efficacy was found to have any predictive value on activity-specific measurements, could explain the relatively
PA behaviour [14]. Thus, the remainder of this section will small increases in self-efficacy noted by the authors. It is clear
Journal of Obesity 3

that more work is needed in the assessment of self-efficacy in the time interval between intention and behavior. Previous
relation to specific behaviors, particularly if the magnitude meta-analytic reviews [24] found that the strength of the
of self-efficacy towards a specific activity is to be accurately intention-behavior association did not decline over time
measured. which contradicts earlier recommendations that intention
In summary, research has found a consistent relationship should be measured as close in time as possible to the behav-
with efficacy and exercise participation in a variety of ior [20]. Recent evidence however suggests that intention-
contexts. The predictive power and ease of operation has behavior relationships do weaken over time [25]. In their
made the self-efficacy theory one of the most consistent recent meta-analytic review, Downs and Hausenblas [26]
predictors of health-related behaviours. What is abundantly did attempt to address this criticism by reviewing 111
clear is that while self-efficacy has been studied extensively, it TRA/TPB studies carried out within the exercise or PA
has usually been incorporated into other behavioral models. domain and specifically examined the predictive utility of the
Researchers have recognised the predictive utility of self- intention-behavior relationship. As the authors predicted,
efficacy and feeling that the predictive power of their own the intention-behavior association was seen to be larger in
theories may be limited, have incorporated self-efficacy. The studies that measured intention and behavior within a 1-
TPB is one such exemplar and will now be discussed. month period compared to the studies with a time interval
greater than 1 month. It seems prudent to suggest that
to improve the predictive utility of the intention-behavior
5. The Theory of Planned Behaviour within association, intention assessment must be measured as close
Physical Activity as possible to the commencement of required behaviour.
Secondly, in their meta-analysis, Hagger et al. [23] found
Identifying the decision making process is at the forefront of that studies with older participants (26 and over) had a
research into increasing PA levels. One such model that has stronger intention-behavior association compared to studies
been used extensively to understand the influencing factors with younger participants (25 and under). Without doubt
of adoption, motivation and adherence to PA is The TPB intentions between children and older individuals may differ
[19]. The development of this theory was built upon previous and this seems to have been overlooked in previous research.
work, The Theory of Reasoned Action (TRA) [20]. The ini- When we consider other subject characteristics such as
tial model proposed that performance of volitional behaviors ethnicity and gender it is apparent that these considerations
(acting without constraints) such as PA is best predicted may have been overlooked by previous authors. For those
from an individual’s stated intention to participate in that reasons it is crucial that further research determines whether
activity. The authors believed that intention is the most these subject characteristics moderate the effect sizes of the
immediate or proximal determinant of behavior influenced TRA and TPB constructs.
by two social cognitive variables, attitude and subjective It has been shown that the application of both the
norm. Despite receiving widespread support in predicating TRA/TPB has strong support in the literature and has aided
intentions and behaviour across a range of health behaviors the understanding of the intention behavior relationship.
including smoking, sexual behavior, and food choice [21], However because of the disparity in the implementation of
the model was expanded upon through the inclusion of the the theory and the methodological and conceptual critiques
perceived behavioral control (PBC) construct. Though the proposed, we are left with many unanswered questions.
theory was initially developed to predict volitional behaviors, While this theory could be used to guide future interventions,
the inclusion of PBC is important as it helps identify personal greater care is warranted regarding measurement of its
and environmental factors not under complete volitional constructs. Nevertheless, it appears that because of the high
control. To summarise, propositions of the TPB include the amount of unexplained variance between intention and
following: (a) individuals will engage in a behavior when they behavior associated with the TPB, researchers have tended
evaluate it positively (attitude), believe that significant others to rely on other theories when developing PA interventions.
want them to engage in it (subjective norm), and perceive it As research is ongoing to enhance the predictive validity
to be under their control (perceived behavioral control); and of the TPB, it is questionable given its ambiguity whether
(b) strong intention and PBC will increase the likelihood of researchers should rely solely upon this theory when con-
a behavior. structing future PA interventions.
Evidence to date has shown that both the TRA and
the TPB perform well in explaining intentions in a range
of populations [22]. Specifically within PA, meta-analytic 5.1. Self-Determination Theory. While some individuals par-
reviews have been conducted and have consistently found ticipate in regular PA simply for the enjoyment of exercising,
the TPB to be superior in the prediction of behavior over others appear to exercise to attain intrinsic or extrinsic
the TRA [23, 24]. While there is certainly abundant evidence rewards such as losing weight, being more attractive, or
supporting the use of TRA/TPB within the PA domain, it has obtaining recognition from significant others [27]. Previous
not been without criticism. research has shown that individuals who exercise out of
Several conceptual and methodological concerns have enjoyment rather than being motivated by intrinsic or exter-
emerged regarding the effectiveness of the TRA and TPB nal rewards are more likely to adhere to a specified exercise
for explaining intention and behavior. Firstly, there seems to programme [28]. Since the goal of health professionals
be discrepancies in the literature regarding measurement of is to promote a continued active lifestyle in individuals
4 Journal of Obesity

not currently meeting the current PA recommendations, basic premise that the most useful theories of motivation
studying the cognitions that are related to motivation has would be broad in scope, encompassing a wide range of
recently been at the centre of much investigation into phenomena, use concepts that have phenomenological or
understanding how to promote long-term behaviour change. personal meaning for people, be derived using empirical
One theory of human motivation that has been applied methods and have principles that can be applied across
extensively to the understanding of exercise behaviour is the life’s domains, [28] the SDT has evolved in the form of
self-determination theory (SDT) [29]. mini-theories. So in a sense the SDT is actually a meta-
Although there are many approaches to initiating theory comprised of subtheories that seek to explain human
behaviour change, research has shown that, without suc- motivation and behaviour based on individual differences.
cessful behavioural intervention, approximately 50% of Rather than being stand-alone theories, the minitheories are
individuals who start a PA program will, on average, drop readily integrateable with one another in that they all share
out within the first six months [30]. The SDT, in contrast, organismic and dialectical assumptions and all involve the
focuses on the processes through which a person acquires concept of basic psychological needs [27]. Thus, together
the motivation for initiating new health-related behaviours they constitute the SDT and, when coordinated, cover all
and maintaining them over time. The theory assumes that types of human behaviour in all domains. We will now briefly
individuals by nature are active, interested, curious, self- discuss the three constituent theories of the SDT.
motivated, and eager to succeed. What it also recognises
though is that individuals can be alienated or passive and
disaffected and accounts for these differences in terms of 5.3. Cognitive Evaluation Theory. The cognitive evaluation
the types of motivation, which stems from the interaction theory (CET) was the first subtheory to be developed to
between individuals inherent active nature and the social explain the effects of intrinsic motivation on behaviours and
environments that either support or thwart that nature [27]. how social contexts affect intrinsic motivation. The theory
Specifically, the SDT proposes that behavioural regula- was formulated to account for reward effects on intrinsic
tion towards an activity varies in the extent to which it motivation and suggests that autonomy and competence
is autonomous (self-determined), which involves behaving are integral constructs of intrinsic motivation and that
with a full sense of volition and choice, or controlling, contextual events, such as rewards, positive feedback or the
which involves behaving with the experience of pressure imposition of a deadline, are likely to have an affect on
and demand toward specific outcomes that comes from an individuals intrinsic motivation towards a behaviour or
forces perceived to be external to the self [28]. Furthermore, activity [29].
the theory proposes that individuals have three basic psy- Research examining the undermining effect and the
chological needs, autonomy, competence and relatedness. informational function of the reward in the CET has been
Autonomy refers to being the perceived origin or source of extensive [32] although very few have actually been carried
ones own behaviour, competence, refers to feeling effective in out in the domain of exercise and health. A possible
ones ongoing interactions with the social environment and explanation for this is that exercise behaviour is unlikely
experiencing opportunities of fulfillment, while relatedness to be viewed as an interesting endeavour and performed
refers to feeling connected to others and to have a feeling solely for extrinsic rewards. This seems plausible as it
of belongingness with individuals and the community [27]. should be remembered that the CET when formulated was
The existence of these needs has been proven empirically [31] only proposed to apply to behaviours and tasks that are
and is perceived to be essential to all individuals as they can highly interesting rather than to mundane tasks that could
act in an intrinsically motivated fashion towards behaviour be viewed as monotonous and boring and unlikely to be
because they perceive it as being important in satisfying these intrinsically motivated [29]. Consequently, the CET cannot
psychological needs. explain how uninteresting behaviours can be prompted or
So when these three basic psychological needs are satis- more importantly how to promote self-regulation of these
fied, an individual’s inherent activity will be supported, opti- behaviours so that individuals will persist over a prolonged
mal motivation will be promoted, and positive psychological, period of time. As initial research utilising the CET focused
developmental, and behavioural outcomes will be produced on the intrinsic-extrinsic distinction, it seems that other
[27]. Conversely, social environments that thwart satisfaction forms of motivation may be apparent in the control of
of these needs yield less optimal forms of motivation and behaviour in uninteresting behaviours and explains why so
have deleterious effects on a wide variety of well-being few studies have been carried out using the CET in the
outcomes. To conclude, it is evident that the SDT is a dialectic exercise and health domain.
theory which views the environment as nurturing need- A further limitation of the CET is that although being
satisfaction and motivation. identified as an important psychological need, relatedness
is not accounted for within the theory [29, 31]. Research
has shown that extrinsic rewards often undermine intrinsic
5.2. Components of the SDT. As an explanation into the motivation, although it is also the case that individuals can
motives behind individuals partaking in, and maintaining, feel autonomous while being motivated. In an attempt to
an active lifestyle through changes in their behaviour, the explain this and incorporate the basic psychological need
SDT has demonstrated exceptional longevity since its initial of relatedness within the SDT, a second subtheory was
conception more than three decades ago. Beginning with the proposed.
Journal of Obesity 5

5.4. Organismic Integration Theory. The Organismic Inte- relatedness in exercise contexts [37]. Given that optimising
gration Theory (OIT) extends the essential distinction participant motivation is a central issue in PA interventions
between intrinsic and extrinsic motivation in the CET and and that the satisfaction of the three basic needs is seen as
seeks to explain the motives behind individuals engaging fundamental to individuals approaching behaviours in an
in nonintrinsically motivated behaviours. Based on the intrinsically motivated fashion, it is perplexing to consider
basic psychological need of relatedness, it is proposed that why this component within the SDT has been poorly
individuals will seek satisfactory relationships with others examined in a PA behavioural context.
and engage in non-intrinsically motivated behaviours to
satisfy relatedness and function effectively in the social world
[27]. Accordingly motivation for the behaviour can vary 5.6. Evaluation and Critique of the Self-Determination The-
depending on the degree to which the value and regulation ory within Physical Activity. Until recently the vast body
of the requested behaviour have been internalized and of research that has adopted tenants of the SDT within
integrated [27]. an exercise domain has not been without its limitations.
Numerous experimental and field studies have examined Much of the research that has been carried out in the
the correlates and consequences of autonomous and con- exercise domain has been methodologically flawed. In their
trolled motivation and have consistently demonstrated that recent meta-analysis Chatzisarantis et al. [36] identified
autonomous regulation is associated with greater persistence; 21 published articles that adopted the Perceived Locus Of
more positive affect; enhanced performance, especially on Control (PLOC) in the exercise domain and found that,
heuristic activities; and greater psychological well-being on the whole, findings supported the existence of a self-
[28]. There have also been studies that have examined determination continuum. Nevertheless, the authors did
the utility of self-determined forms of motivation within question this assumption due to the small number of studies
an exercise context, demonstrating that autonomous moti- within the meta-analysis and the distinct lack of consistency
vation is associated with exercise behavioural engagement in the methodological approach used within the studies.
and adherence over time [33], exercise intentions [34], Much of the studies seemed to adopt correlational designs
and perceived competence [35]. These findings have been rather than experimental designs. Given that researchers
supported through a recent meta-analysis of 21 studies which can only infer causality when adopting correlational designs
confirmed the simplex-ordered structure of relations among it is reasonable to agree with the tentative findings of
the regulation styles and the effects of these on exercise Chatzisarantis et al. [36] specifically with regards to the
behaviours and outcomes [36]. process of internalization.
While experimental methods and intervention studies
have been extensively adopted by researchers in the sporting
5.5. Basic Needs Theory. The final subtheory within the domain, until recently very little research had adopted
SDT refers to the basic needs theory. This last theory was these approaches within an exercise behavioural context.
only recently developed in an attempt by Deci and Ryan, Edmunds et al. [38] though have shown that interventions
[27] to clarify the somewhat often misunderstood meaning designed to change motivational regulations in individuals
of the basic psychological needs of individuals and their can produce increases in exercise behaviour. Specifically,
relationship to mental health and well-being. As highlighted the researchers examined whether an exercise instructor’s
previously the three basic needs refer to autonomy, com- teaching style could be manipulated so that it is was per-
petence, and relatedness. The basic needs theory proposes ceived by individuals as providing more autonomy support,
that when individuals feel as though these three needs are structure, and interpersonal involvement. In addition the
supported, both intrinsic motivation and internalization authors also examined the impact of the delivery of the
are supported. Conversely, if the social context inhibits or exercise class on the basic psychological needs, autonomous
neglects one of these needs then intrinsic motivation and motivation, and behavioural outcomes. Of the two groups
internalization will be reduced. So, within an exercise-related that participated in the study, one group acted as a control
context, if individuals feel as if a particular behaviour has group whereas in the second group, the treatment group, the
the potential to fulfil valued goals then participation in that instructor focused upon promoting autonomy support by
behaviour will increase. With such endeavours the authors taking the perspective of the exercise class participants into
propose that the behaviours will become internalised, finally account, acknowledging their feelings and providing them
becoming integrated into an individuals set of behaviours with pertinent information and opportunities for choice.
that will satisfy the three basic psychological needs. In comparison to the control group, the individuals
Understanding the conditions that foster, rather than within the treatment group reported significantly greater
undermine, these psychological needs has enormous poten- increases in relatedness and competence need satisfaction,
tial for the development of social environments that will positive affect, structure, and interpersonal involvement
promote self-determined motivation, personal development, which corresponded to increased participation and retention
and well-being. Although it is well documented that the rates. The authors concluded that interventions grounded in
satisfaction of the needs for autonomy, competence, and the SDT could positively influence exercise class participants’
relatedness will predict health and well-being [27, 28], behavioural, cognitive, and affective responses to exercise
surprisingly little research has been carried out on the and should be considered when designing future interven-
measurement of perceived competence, autonomy, and tions studies. Unfortunately, and as the authors state, this
6 Journal of Obesity

study was not without its limitations. Firstly, the investiga- that each stage is characterised by particular cognitive
tion only lasted 10 weeks and failed to consider whether this and psychodynamic variables, which, if targeted, can move
approach would lead to long-term participation. Secondly, individuals into subsequent stages of behavior. The processes
a lack of generalizability of the findings is present since of change included in the TTM represents the behavioral or
only female participants aging between 18 and 53 years were experiential changes used by an individual to modify their
recruited. experiences.
Although numerous motivational theories exist, re- The body of evidence regarding the effectiveness of TTM
searchers seeking to understand the social conditions that based interventions is mixed. In relation to PA, van Sluijs
support need satisfaction and subsequent motivation have et al. [44] carried out a systematic review of the literature
shown a keen interest in the application of the SDT within concerning the effect of stages of change-based interventions
health-related exercise contexts. The theoretical framework and its effects on smoking, PA, and dietary behavior. In
proposed has been well supported within a variety of relation to PA the authors identified 13 randomised control
contexts, including PA, and is increasingly being used and trials (RCTs) that included stage of change-based and behav-
recommended to guide future intervention and experimental ioral outcomes and found no evidence of an advantageous
studies. While recent research has supported the application effect of stage-based interventions as opposed to alternative
of interventions grounded in SDT [38], it is evident from approaches. This reflects the findings of previous systematic
this discussion that further research is needed within the reviews [45].
field of exercise-related behaviour change. The theory has Here the authors systematically assessed the effective-
much to offer in terms of predicting behaviour, understand- ness of PA promotional activities that used a stage-based
ing behavioural mechanisms, and designing appropriate approach in bringing about changes in health-related behav-
interventions that increase PA participation and adherence. ior. Their inclusion criteria of only using RCTs identified
Encouraging individuals to self-regulate and continually seven trials of activity promotion interventions based on
form intentions to exercise seem to be the best option to the constructs of the TTM. Four of these seven studies
promote adherence to a physically active lifestyle. found no significant changes in behavioral outcomes, two
studies showed mixed results, while only one study showed
significant effects in favour of stage-based interventions.
6. The Transtheoretical Model within The authors did however raise issue with the methodolog-
Physical Activity ical quality of the studies used highlighting the lack of
consistency between interventions and poor appreciation
In an attempt to overcome the limitations of social cognitive of participant stage classification. The authors deemed the
models, researchers have considered the use of stage-based TTM an unsuitable approach for bringing about positive
approaches in behavior change interventions. The most changes in health behavior though they did suggest the need
popular stage model applied with a PA context is the for further research.
TTM [39]. Within the TTM health behavior adoption Finally, Adams and White [46] reviewed the effectiveness
and maintenance is described as a cyclic process whereby of 16 TTM informed interventions and found that 73 per
individuals pass through a series of specific stages, each cent of short-term (<6 month) studies reported a positive
characterised by a particular pattern of psychosocial and effect of TTM studies over “control conditions.” The equiv-
behavioral changes. As such, within the TTM, individuals alent long-term (>6 months) proportion was 29 per cent.
are classified by their readiness to change into one of From these findings it could be suggested that stage-based
five stages: precontemplation, contemplation, preparation, interventions are no more effective than control conditions
action, and maintenance [40]. Though several attempts in promoting long-term adherence to PA. However, Marshall
at change are likely before maintenance is reached, the and Biddle [47] in their meta-analysis included 71 published
progression through the process may in fact strengthen reports that presented empirical data on at least one core
behavior change as individuals learn from past regressions construct of the TTM applied to exercise and PA. From their
[41]. analysis they were able to support the application of the TTM
Designing interventions that meet the requirements of suggesting that the core constructs of the TTM differ across
individuals based on their stage of change is considered stages with most in the direction predicted by the theory.
to not only enhance participation and retentions rates but There may be numerous reasons why stage-based inter-
also reduce the resistance of individuals to initiate difficult ventions may sometimes lack effectiveness though three
behaviors [42]. Consequently, much of the PA research that dominant explanations have been suggested. First, a range
has utilised the TTM in interventions has been delivered of evidence has been used not only to evaluate the model but
in accordance to an individual’s stage of change and have also to develop arguments in the literature [48]. For instance,
found that individuals progress more towards the action and there seems to be no evidence-based consensus on which
maintenance stages than control participants [43]. Another criteria should be used for assessing the methodological
valuable feature of the TTM is that not only can it provide quality of studies. Whilst some research has utilized an RCT
a framework to categorise individuals into a particular stage design, others have no control group with much being cross-
of change, but it also indicates how to encourage individuals sectional in nature [49]. Consequently, the items included
to change their behaviour and progress through the different are, to some extent, arbitrarily chosen. Deciding upon which
stages. Specifically, Prochaska and Marcus [42] propose studies are included and excluded in systematic reviews
Journal of Obesity 7

therefore can significantly influence the evidence gathered considered when planning and implementing health pro-
and thus the consensus given. motion interventions [8]. When adopting such an approach
Secondly, since there are fundamental differences be- it is acknowledged that behavior is influenced by multiple
tween some health behaviours and the addictive behaviors levels and, in order to ensure lasting change, appreciation of
upon which the model was originally formulated, a lack this is necessary. The use of multilevel ecological approaches
of evidence may be due to the fact that some behaviors is widely accepted to guide public health policy in the
are simply more suitable to stage-based interventions [48]. United States (US) [53]. Though the use of ecological
Given that PA is a multifaceted complex phenomenon, approaches within health promotion is relatively recent, its
practitioners who are reliant upon models that focus application for understanding behavior is not new. The role
upon just the individual may be underestimating the true of the primary contributors to ecological perspectives and its
complexity of influences upon the individual. There is a subsequent utility within health research has been identified
wealth of evidence that has shown that other external and discussed previously [8, 54, 55] and will not be revisited
and social factors, such as age, gender, and socioeconomic here. Rather, we will identify the challenges of researchers
status, influence complex behaviors, which the TTM fails to embracing an ecological perspective to inform interventions
consider [6, 8, 50]. It seems that the reliance upon the TTM within the domain of PA behavior.
by practitioners helps conceal the multifaceted complexities Though ecological-based multilevel interventions hold
involved in changing complex behaviors. great potential for influencing PA, certain issues associated
with such an approach need to be acknowledged. For
instance, guiding frameworks such as the Social Ecological
7. Emerging Practice: An Ecological Approach Theory [8] and the Ecological Model of Health Behavior
Within this paper, only a small selection of behavior theories [50] are useful for considering the behavioral determinants
has been discussed. The inclusion of the models in this paper of health broadly. However, these models were devised to
is based on their popularity and application in intervention provide an overarching framework to guide interventions
studies within the PA domain. While all four models have and their role in advancing the application of ecological
various contrasting features, they also share some core perspectives within health behavior should be commended.
principles. First, all of the models relate predominantly on Nevertheless, these frameworks fail to provide specific mech-
changing the behavior of the individual and focus less on anisms through which particular influences may interact and
the environment. Second, the models exist within a positivist influence specific behaviors. Indeed, the lack of specificity
and cognitive-rational paradigm where the main focus is to and instruction given presents methodological and concep-
predict and control. As such, the models consider the deter- tual challenges that are not apparent when utilising cognitive
minants of specific behaviors as linear and maybe even more based models.
importantly phase staged. Meanwhile, Adams and White [51] Understandably, it could be for this very reason that
state that individualized stage-based interventions are not researchers within the PA domain are still heavily reliant
effective in promoting long-term adherence to PA as they upon cognitive-based theories which provide applicable
oversimplify the individual’s ability to make positive changes measures that can be implemented across numerous
to their behavior. domains and setting, particularly when one considers the
Overall, traditional health behavior models leave a sug- positivist reductionist approach which seems to govern
gestion that interventions, programs, activities, and policies funding bodies and public health policy which advocates
can be fully planned and controlled, with predictable out- the need for identifying causal pathways in health research.
comes. This clearly does not reflect the real world of health When one considers further the traditional desire of inferring
behavior and PA seen from the developments in obesity causality and solving problems through rational deduction
and overweight rates throughout the world. What has now it is unsurprising that there is still reliance from researchers
become apparent within the last twenty years is that changing upon traditional linear cognitive paradigms to inform inter-
health behavior is a complex and multifaceted phenomenon ventions.
that has multiple levels of influences [8, 52]. With the With such an approach there is a belief that behavior
reliance upon individual psychosocial models it is now clear can be explained as a linear process whereby decisions are
that such approaches do little to effect change beyond the planned and actions instinctively ensue. It is apparent that
individual. Since social ecological models of health behavior complex behaviors such as PA do not occur in such a way.
[8] focus on individual influences as well as social, policy, Instead, behavior is influenced by multiple levels of factors
and environmental factors that may facilitate or inhibit that interact with one another influencing individuals and
individual behaviour, researchers have now embraced the use subsequent behavior. As such it is now becoming accepted
of such frameworks to inform their interventions. that behavior cannot be understood by measuring individual
factors alone but rather behavior emerges due to the complex
interactions between multiple levels of influences [55–57].
8. Ecological Models of Physical Increasing acceptance of the complexities involved in
Activity Promotion human behavior renders the current linear phase stage
approach to understanding PA behaviors incongruous. From
Ecological models profess that individual, interpersonal, a PA perspective, the complexities involved in understand-
organizational, societal, and community factors should be ing behavior can be viewed as a collection of numerous
8 Journal of Obesity

determinants whose actions, though unpredictable, impact collection of numerous determinants whose actions, though
upon other determinants [6]. Since these determinants unpredictable, impact upon other determinants. Since these
of behavior are often nested within numerous levels of determinants of behaviour are often nested within numerous
influences as outlined by the ecological framework [8], to levels of influences as outlined by the ecological framework,
fully comprehend the impact of individual determinants to fully comprehend the impact of individual determinants
upon behavior there must be an acceptance of the mediating upon behavior there must be an acceptance of the mediating
role of all determinants. Another important premise to role of all determinants [62].
acknowledge is that influence of determinants may change Despite gathering consensus that multilevel approaches
over time with a direct influence upon behavior [58]. This are required to ensure long-lasting change in behaviour, the
is an important constituent to the study of human behavior field is limited at present due to the dependence upon cross-
which is not embraced within the framework of the four sectional evidence [63]. Of course this new approach to the
main theoretical models most widely utilized within the PA study of health behavior is in its infancy and as one would
domain. expect there is still a great deal to understand regarding the
Rather, the main determinants are believed to influence key determinants of behavior worthy of intervening upon.
individuals whereby behavior change occurs in a linear Nevertheless, the PA domain has begun to embrace the
manner [59]. For instance, traditional research paradigms need for ecologically informed PA interventions [64–66].
propose that the measurement, and increases, of cognitions It could be that we are at the embryonic stage of another
such as attitudes, efficacy, beliefs and intentions over time paradigm shift with the acceptance that behavior change
will determine how effective an intervention is for enhancing does not occur in a deterministic and linear fashion, but
behavior. This approach however seems flawed as it fails more so through complex interactions between numerous
to appreciate the complexities involved in human behavior determinants interacting in a nonlinear manner [59, 61].
whereby change often occurs in a nonlinear manner [59, 60]. This principle effectively precludes the inference of
causality derived from cross sectional analysis. This premise
though is not new given that Bronfenbrenner [67] identified
9. Chaos and Complexity these issues more than 60 years ago stating “piecemeal
analysis, fixed in time and space, of isolated aspects and
Nonlinear dynamical approaches, such as complexity theory attributes is insufficient and even misleading” [67]. Non-
and chaos theory, to the study of health behaviours have linear dynamical approaches, such as complexity and chaos
recently emerged in the health domain literature [57, 59– theory, to the study of health behaviors have recently
61]. As these authors contend, this new approach to emerged in the health domain literature [57, 59–61]. The
understanding behavior has stemmed from the limitations proposed application of the new science of complex adaptive
inherent in current approaches that are reliant solely upon systems within the PA domain may however provide a means
traditional cognitive rationale paradigms to explain complex to better understand the complexities involved in human
behaviours. With such an approach there is a belief that behavior. Within a complex adaptive system approach
behaviour can be explained as a linear process whereby human beings can be viewed as being composed of, and
decisions are planned and actions instinctively ensue. From operating within, multiple interacting and self adjusting
the previous discussion however, it is apparent that complex systems including individual, interpersonal, organizational,
behaviours such as PA do not occur in such a way. societal, and community systems [60]. From a PA per-
Instead, behavior is influenced by multiple levels of factors spective, this complex adaptive system can be viewed as a
that interact with one another influencing individuals and collection of numerous determinants whose actions, though
subsequent behaviour, hence the term, complex behaviours. unpredictable, impact upon other determinants. Since these
As such it is now becoming accepted that behavior cannot be determinants of behaviour are often nested within numerous
understood by measuring individual factors alone but rather levels of influences as outlined by the ecological framework,
behavior emerges due to the complex interactions between to fully comprehend the impact of individual determinants
multiple levels of influences which is often spontaneous, upon behavior there must be an acceptance of the mediating
uncontrolled, and uncertain [56], hence the term chaotic. role of all determinants [62].
Increasing acceptance of the complexities involved in Another important premise of complex adaptive sys-
human behavior, for instance, the initiation and/or main- tems is that determinants can change which can affect
tenance of PA, renders the current linear phase stage the behavior of individuals over time [57]. This is an
approach to understanding PA incongruous. The proposed important constituent to the study of human behavior and
application of the new science of complex adaptive systems highlights the limitations of current research practices within
with the PA domain [59, 60] may however provide a means the PA domain. For instance, despite gathering consensus
to better understand the complexities involved in human that multilevel approaches to health behavior consistent
behavior. Thus, within a complex adaptive system approach with social ecological frameworks are required to ensure
human beings can be viewed as being composed of, and substantial changes in health behaviors, the field is limited at
operating within, multiple interacting and self-adjusting present due to the dependence upon cross sectional evidence
systems including individual, interpersonal, organizational, [63]. In complex systems, behavior is said to emerge from the
societal, and community systems [8, 50, 57]. From a PA interaction over time of numerous determinants influencing
perspective, this complex adaptive system can be viewed as a said behavior in a non linear manner [61]. Traditional
Journal of Obesity 9

research paradigms propose that the measurement, and current linear phase stage approach to understanding PA
increases, of cognitions such as attitudes, efficacy, beliefs, and behaviours incongruous. What has now become apparent
intentions over time will determine how effective an inter- within the last twenty years is that changing health behavior
vention is for enhancing health behaviors. This approach is a complex and multifaceted phenomenon that has multiple
however seems flawed as it fails to appreciate the complexities levels of influences [8, 52]. Since ecological approaches
involved in human behavior whereby change often occurs towards health behavior focus on individual influences as
in a nonlinear manner, rather than the linear, deterministic well as social, policy researchers have now embraced the use
manner proposed by traditional cognitive paradigms [60]. of such frameworks to inform their interventions. However,
The seminal paper by Resnicow and Vaughan [59] has there is a lack of models available that provide specific
ignited debate in the literature [68] regarding the usefulness mechanisms through which particular influences may inter-
of the prevailing health behavior theories in understanding act and influence behavior. Researchers are now faced with
PA behavior which warrants further comment. Despite the daunting task of relying upon detailed cross sectional
disagreements between authors, both contend that there evidence to develop and test hypothesis in order to enhance
is a need for further research that incorporates nonlinear further our understanding of the determinants of PA. It
concepts into future interventions. While this sentiment seems that the field of PA research is at the embryonic stage
is shared there is at present a lack of guidance upon of a paradigm shift towards improving our understanding
how to implement such approaches within a PA context. of complex behaviours through the application of complex
Nevertheless, previous and future reviews based upon cross ecological interventions. Recently the health domain has
sectional research designs can help inform interventions and embraced the use of nonlinear dynamical approaches, such
generate hypotheses and their importance should not be as complexity theory to the study of complex behaviors. In
underestimated. the PA domain, there is no unified model of research and
practice which integrates both ecological and complexity
theories. Forthcoming PA models should begin to consider
10. Summary the advantages of incorporating the principles of complexity
theories into future intervention programmes. This domain
Excellent reviews have already taken place which has en-
of future work promises to be a worthwhile endeavour in
hanced our understanding of the key determinants of
tackling current obesity and inactivity levels and should help
specific behaviors, within specific contexts [52, 69–72].
generate further understanding of the complexities involved
Nevertheless, researchers are faced with the difficult task
in PA behavior.
of developing site- and behavior-specific ecological models.
Though daunting the study of human behavior through
ecological and complexity theories affords the generation References
of hypothesis informed through cross sectional evidence.
Identifying the key determinants of behavior relative to [1] World Health Organization, “A healthy city is an active city: a
individual characteristics, contexts, and activities is the first physical activity planning guide,” 2008.
[2] S. N. Blair and J. N. Morris, “Healthy hearts—and the
step in devising appropriate interventions. The challenge
universal benefits of being physically active: physical activity
thereafter is to consider how to monitor the complex and health,” Annals of Epidemiology, vol. 19, no. 4, pp. 253–
interactions that will occur over time within the multiple 256, 2009.
levels of influence. As such, a unified model of research [3] J. F. Sallis, R. B. Cervero, W. Ascher, K. A. Henderson, M. K.
and practice which integrates both ecological and complexity Kraft, and J. Kerr, “An ecological approach to creating active
theories is very much needed. Only then can we begin to living communities,” Annual Review of Public Health, vol. 27,
understand the role of the multiple influences upon behavior pp. 297–322, 2006.
and begin to translate this evidence into future health- [4] S. Sutton, “How does the Health Action Process Approach
enhancing interventions. (HAPA) bridge the intention-behavior gap? An examination
of the model’s causal structure,” Applied Psychology, vol. 57,
no. 1, pp. 66–74, 2008.
11. Conclusion [5] C. R. Nigg, B. Borrelli, J. Maddock, and R. K. Dishman, “A
theory of physical activity maintenance,” Applied Psychology,
It is now well established that theoretically informed in- vol. 57, no. 4, pp. 544–560, 2008.
terventions are imperative for successful physical activity [6] J. F. Sallis, N. Owen, and E. B. Fisher, “Ecological models of
promotion. Though only a small selection of behavior health behavior,” in Health Behavior and Health Education:
theories have been discussed in this paper it is apparent Theory, Research and Practice, K. Glanz, B. K. Rimer, and
that two major approaches to PA promotion have dominated K Viswanath, Eds., pp. 465–485, Jossey-Bass, San Francisco,
Calif, USA, 4th edition, 2008.
the literature, one founded on a stage-based model and the
[7] N. Humpel, N. Owen, and E. Leslie, “Environmental factors
other founded on social cognitive principles. While these associated with adults’ participation in physical activity. A
approaches have greatly enhanced our understanding of review,” American Journal of Preventive Medicine, vol. 22, no.
the key determinants of PA behavior it is now apparent 3, pp. 188–199, 2002.
that behavior is influenced not only through individual [8] D. Stokols, “Social ecology and behavioral medicine: implica-
level cognitions. It seems that the increasing acceptance of tions for training, practice, and policy,” Behavioral Medicine,
the complexities involved in human behavior renders the vol. 26, no. 3, pp. 129–138, 2000.
10 Journal of Obesity

[9] T. K. Scanlan, P. J. Carpenter, G. W. Schmidt, G. W. Schmidt, the context of physical activity,” Journal of Sports Sciences, vol.
J. Simons, and B. Keeler, “An introduction to the sport 23, no. 1, pp. 49–61, 2005.
commitment model,” Journal of Sport & Exercise Psychology, [26] D. S. Downs and H. A. Hausenblas, “The theories of reasoned
vol. 15, no. 1, pp. 1–15, 1993. action and planned behavior applied to exercise: a meta-
[10] C. J. Browning and S. A. Thomas, “Models of behaviour analytic update,” Journal of Physical Activity and Health, vol.
change in health promotion,” in An Evidence-Based Handbook 2, no. 1, p. 76, 2005.
for Social and Public Health, C. J. Browning and S. A. Thomas, [27] E. L. Deci and R. M. Ryan, “The “what” and “why” of goal
Eds., pp. 7–15, Elsevier, Scotland, UK, 2005. pursuits: human needs and the self-determination of behav-
[11] A. Bandura, Social Learning Theory, Prentice Hall, Englewood ior,” Psychological Inquiry, vol. 11, no. 4, pp. 227–268, 2000.
Cliffs, NJ, USA, 1977. [28] E. L. Deci and R. M. Ryan, “Facilitating optimal motivation
[12] E. McAuley, G. J. Jerome, D. X. Marquez, S. Elavsky, and and psychological well-being across life’s domains,” Canadian
B. Blissmer, “Exercise self-efficacy in older adults: social, Psychology, vol. 49, no. 1, pp. 14–23, 2008.
affective, and behavioral influences,” Annals of Behavioral [29] E. L. Deci and R. M. Ryan, Intrinsic Motivation and Self-
Medicine, vol. 25, no. 1, pp. 1–7, 2003. Determination in Human Behavior, Plenum Press, New York,
[13] B. Resnick, M. H. Palmer, L. S. Jenkins, and A. M. Spellbring, NY, USA, 1985.
“Path analysis of efficacy expectations and exercise behaviour [30] R. K. Dishman and J. Buckworth, “Increasing physical activity:
in older adults,” Journal of Advanced Nursing, vol. 31, no. 6, pp. a quantitative synthesis,” Medicine and Science in Sports and
1309–1315, 2000. Exercise, vol. 28, no. 6, pp. 706–719, 1996.
[14] R. K. Dishman, R. W. Motl, R. Saunders et al., “Self-efficacy [31] K. M. Sheldon, A. J. Elliot, Y. Kim, and T. Kasser, “What is
partially mediates the effect of a school-based physical-activity satisfying about satisfying events? Testing 10 candidate psy-
intervention among adolescent girls,” Preventive Medicine, vol. chological needs,” Journal of Personality and Social Psychology,
38, no. 5, pp. 628–636, 2004. vol. 80, no. 2, pp. 325–339, 2001.
[15] M. S. Hagger and N. Chatzisarantis, Social Psychology of [32] E. L. Deci, R. Koestner, and R. M. Ryan, “A meta-analytic
Exercise and Sport, Open University Press, Maidenhead, UK, review of experiments examining the effects of extrinsic
2005. rewards on intrinsic motivation,” Psychological Bulletin, vol.
[16] E. McAuley, B. Blissmer, J. Katula, and T. E. Duncan, “Exercise 125, no. 6, pp. 627–668, 1999.
environment, self-efficacy, and affective responses to acute [33] N. L. D. Chatzisarantis, M. S. Hagger, S. J. H. Biddle, and
exercise in older adults,” Psychology and Health, vol. 15, no. C. Karageorghis, “The cognitive processes by which perceived
3, pp. 341–355, 2000. locus of causality predicts participation in physical activity,”
[17] E. McAuley, B. Blissmer, J. Katula, T. E. Duncan, and S. Journal of Health Psychology, vol. 7, no. 6, pp. 685–699, 2002.
L. Mihalko, “Physical activity, self-esteem, and self-efficacy [34] M. S. Hagger, N. L. D. Chatzisarantis, V. Barkoukis et al.,
relationships in older adults: a randomized controlled trial,” “Cross-cultural generalizability of the theory of planned
Annals of Behavioral Medicine, vol. 22, no. 2, pp. 131–139, behavior among young people in a physical activity context,”
2000. Journal of Sport and Exercise Psychology, vol. 29, no. 1, pp. 2–
[18] C. B. Dallow and J. Anderson, “Using self-efficacy and a trans- 20, 2007.
theoretical model to develop a physical activity intervention [35] M. Goudas, S. J. Biddle, and K. Fox, “Perceived locus of cau-
for obese women,” American Journal of Health Promotion, vol. sality, goal orientations, and perceived competence in school
17, no. 6, pp. 373–381, 2003. physical education classes,” The British Journal of Educational
[19] I. Ajzen, “The theory of planned behavior,” Organizational Psychology, vol. 64, pp. 453–463, 1994.
Behavior and Human Decision Processes, vol. 50, no. 2, pp. 179– [36] N. L. D. Chatzisarantis, M. S. Hagger, S. J. H. Biddle, B. Smith,
211, 1991. and J. C. K. Wang, “A meta-analysis of perceived locus of
[20] I. Ajzen and M. Fishbein, Understanding Attitudes and Predict- causality in exercise, sport, and physical education contexts,”
ing Social Behavior, Prentice Hall, Englewood Cliffs, NJ, USA, Journal of Sport and Exercise Psychology, vol. 25, no. 3, pp. 284–
1980. 306, 2003.
[21] C. J. Armitage and M. Conner, “Efficacy of the theory of [37] R. J. Vallerand and G. F. Losier, “An integrative analysis of
planned behaviour: a meta-analytic review,” British Journal of intrinsic and extrinsic motivation in sport,” Journal of Applied
Social Psychology, vol. 40, part 4, pp. 471–499, 2001. Sport Psychology, vol. 11, no. 1, pp. 142–169, 1999.
[22] N. L. D. Chatzisarantis, C. Frederick, S. J. H. Biddle, M. S. [38] J. Edmunds, N. Ntoumanis, and J. L. Duda, “Testing a self-
Hagger, and B. Smith, “Influences of volitional and forced determination theory-based teaching style intervention in the
intentions on physical activity and effort within the theory of exercise domain,” European Journal of Social Psychology, vol.
planned behaviour,” Journal of Sports Sciences, vol. 25, no. 6, 38, no. 2, pp. 375–388, 2008.
pp. 699–709, 2007. [39] J. O. Prochaska and C. C. DiClemente, The Transtheoretical
[23] M. S. Hagger, N. L. Chatzisarantis, and S. J. Biddle, “A Approach: Crossing Traditional Boundaries of Therapy, Dow
meta-analytic review of the theories of reasoned action and Jones-Irwin, Homewood, Ill, USA, 1984.
planned behavior in physical activity: predictive validity and [40] B. H. Marcus and L. Forsyth, Motivating People to be Physically
the contribution of additional variables,” Journal of Sport and Active, Human Kinetics Books, Champaign, Ill, USA, 2003.
Exercise Psychology, vol. 24, pp. 3–32, 2002. [41] B. H. Marcus and L. R. Simkin, “The stages of exercise
[24] H. A. Hausenblas, A. V. Carron, and D. E. Mack, “Application behavior,” Journal of Sports Medicine and Physical Fitness, vol.
of the theories of reasoned action and planned behavior 33, no. 1, pp. 83–88, 1993.
to exercise behavior: a meta-analysis,” Journal of Sport and [42] J. O. Prochaska and B. Marcus, “The transtheoretical model:
Exercise Psychology, vol. 19, no. 1, pp. 36–51, 1997. application to exercise,” in Advances in Exercise Adherence,
[25] N. L. D. Chatzisarantis, M. S. Hagger, S. J. H. Biddle, and B. R. K. Dishman, Ed., pp. 161–180, Human Kinetics Books,
Smith, “The stability of the attitude-intention relationship in Champaign, Ill, USA, 1994.
Journal of Obesity 11

[43] A. F. Kirk, N. Mutrie, P. D. MacIntyre, and M. B. Fisher, [61] P. E. Plsek and T. Greenhalgh, “Complexity science: the
“Promoting and maintaining physical activity in people with challenge of complexity in health care,” British Medical
type 2 diabetes,” American Journal of Preventive Medicine, vol. Journal, vol. 323, no. 7313, pp. 625–628, 2001.
27, no. 4, pp. 289–296, 2004. [62] A. E. Bauman, “The physical environment and physical
[44] E. M. F. van Sluijs, M. N. M. van Poppel, J. W. R. Twisk, and activity: moving from ecological associations to intervention
W. van Mechelen, “Physical activity measurements affected evidence,” Journal of Epidemiology and Community Health, vol.
participants’ behavior in a randomized controlled trial,” 59, no. 7, pp. 535–536, 2005.
Journal of Clinical Epidemiology, vol. 59, no. 4, pp. 404–411, [63] J. P. Elder, L. Lytle, J. F. Sallis et al., “A description of the social-
2006. ecological framework used in the trial of activity for adolescent
[45] R. P. Riemsma, J. Pattenden, C. Bridle et al., “A systematic girls (TAAG),” Health Education Research, vol. 22, no. 2, pp.
review of the effectiveness of interventions based on a stages- 155–165, 2007.
of-change approach to promote individual behaviour change,” [64] P. J. Naylor, H. M. Macdonald, J. A. Zebedee, K. E. Reed,
Health Technology Assessment, vol. 6, no. 24, pp. 1–231, 2002. and H. A. McKay, “Lessons learned from Action Schools!
[46] J. Adams and M. White, “Are activity promotion interventions BC—an “active school” model to promote physical activity in
based on the transtheoretical model effective? A critical elementary schools,” Journal of Science and Medicine in Sport,
review,” British Journal of Sports Medicine, vol. 37, no. 2, pp. vol. 9, no. 5, pp. 413–423, 2006.
106–114, 2003. [65] D. S. Buchan, S. Ollis, N. E. Thomas, and J. S. Baker, “The
[47] S. J. Marshall and S. J. H. Biddle, “The transtheoretical model influence of a high intensity physical activity intervention on a
of behavior change: a meta-analysis of applications to physical selection of health related outcomes: an ecological approach,”
activity and exercise,” Annals of Behavioral Medicine, vol. 23, BMC Public Health, vol. 10, no. 1, article 8, 2010.
no. 4, pp. 229–246, 2001. [66] P. J. Naylor and H. A. McKay, “Prevention in the first place:
[48] C. Bridle, R. P. Riemsma, J. Pattenden et al., “Systematic review schools a setting for action on physical inactivity,” British
of the effectiveness of health behavior interventions based on Journal of Sports Medicine, vol. 43, no. 1, pp. 10–13, 2009.
the transtheoretical model,” Psychology and Health, vol. 20, no. [67] U. Bronfenbrenner, “A constant frame of reference for
3, pp. 283–301, 2005. sociometric research: part II. Experiment and inference,”
[49] R. West, “Time for a change: putting the transtheoretical Sociometry, vol. 7, pp. 40–75, 1944.
(Stages of Change) model to rest,” Addiction, vol. 100, no. 8, [68] J. Brug, “Order is needed to promote linear or quantum
pp. 1036–1039, 2005. changes in nutrition and physical activity behaviors: a reaction
[50] K. R. McLeroy, D. Bibeau, A. Steckler, and K. Glanz, “An to “a chaotic view of behavior change” by Resnicow and
ecological perspective on health promotion programs,” Health Vaughan,” International Journal of Behavioral Nutrition and
Education Quarterly, vol. 15, no. 4, pp. 351–377, 1988. Physical Activity, vol. 3, article 29, 2006.
[69] J. F. Sallis, J. J. Prochaska, and W. C. Taylor, “A review of
[51] J. Adams and M. White, “Why don’t stage-based activity
correlates of physical activity of children and adolescents,”
promotion interventions work?” Health Education Research,
Medicine and Science in Sports and Exercise, vol. 32, no. 5, pp.
vol. 20, no. 2, pp. 237–243, 2005.
963–975, 2000.
[52] J. F. Sallis and K. Glanz, “The role of built environments in
[70] S. G. Trost, N. Owen, A. E. Bauman, J. F. Sallis, and W. Brown,
physical activity, eating, and obesity in childhood,” The Future
“Correlates of adults’ participation in physical activity: review
of Children, vol. 16, no. 1, pp. 89–108, 2006.
and update,” Medicine and Science in Sports and Exercise, vol.
[53] US Department of Health and Human Services, Healthy 34, no. 12, pp. 1996–2001, 2002.
People 2010, US Department of Health and Human Services, [71] B. E. Saelens, J. F. Sallis, and L. D. Frank, “Environmental
Wsahington, DC, USA, 2000. correlates of walking and cycling: findings from the trans-
[54] J. F. Sallis and N. Owen, “Ecological models,” in Health portation, urban design, and planning literatures,” Annals of
Behavior and Health Education: Theory, Research and Practice, Behavioral Medicine, vol. 25, no. 2, pp. 80–91, 2003.
K. Glanz, F. M. Lewis, and B. K. Rimer, Eds., pp. 304–424, [72] N. W. Burton, G. Turrell, B. Oldenburg, and J. F. Sallis,
Jossey-Bass, San Francisco, Calif, USA, 1997. “The relative contributions of psychological, social, and
[55] L. McLaren and P. Hawe, “Ecological perspectives in health environmental variables to explain participation in walking,
research,” Journal of Epidemiology and Community Health, vol. moderate-, and vigorous-intensity leisure-time physical activ-
59, no. 1, pp. 6–14, 2005. ity,” Journal of Physical Activity and Health, vol. 2, no. 2, pp.
[56] U. Bronfenbrenner, Making Human Beings Human : Bioeco- 181–196, 2005.
logical Perspectives on Human Development, Sage, Thousand
Oaks, Calif, USA, 2005.
[57] T. Wilson, T. Holt, and T. Greenhalgh, “Complexity science:
complexity and clinical care,” British Medical Journal, vol. 323,
no. 7314, pp. 685–688, 2001.
[58] P. E. Plsek and T. Wilson, “Complexity science: complexity,
leadership, and management in healthcare organisations,”
British Medical Journal, vol. 323, no. 7315, pp. 746–749, 2001.
[59] K. Resnicow and R. Vaughan, “A chaotic view of behavior
change: a quantum leap for health promotion,” International
Journal of Behavioral Nutrition and Physical Activity, vol. 3,
article 25, 2006.
[60] K. Resnicow and S. E. Page, “Embracing chaos and complexity:
a quantum change for public health,” American Journal of
Public Health, vol. 98, no. 8, pp. 1382–1389, 2008.