Weighing the odds: an exploration of resistance to obesity and overweight

Denise Conroy (University of Auckland Business School, University of Auckland, Auckland, New Zealand)
Sandra D. Smith (University of Auckland Business School, University of Auckland, Auckland, New Zealand)
Catherine Frethey-Bentham (University of Auckland Business School, University of Auckland, Auckland, New Zealand)

Journal of Social Marketing

ISSN: 2042-6763

Publication date: 8 October 2018



In 2018, we have surpassed the population landmark of 7.5 billion, and yesterday’s global crisis of under-nutrition in poorer nations is now accompanied by a journey into overweight and obesity. The purpose of our research is to focus on the health and resistance of those who avoid overweight and obesity rather than continuing to focus on the pathology and disease of this phenomenon.


Taking a consumer-centric perspective and using the lens of the social-economic framework, the authors report qualitative research conducted with 31 young people (ages 17-26) who have been resistant to weight gain in an increasingly obesogenic environment, followed by a survey of the general population, n = 921. The authors look at this type of consumer resistance to better understand how to develop government and community leadership and build more obesogenically resilient societies.


The findings support the contention that obesity is a social problem that requires a social solution.


The main contribution to the conversation addressing increasing levels of overweight and obesity is that this research demonstrates that these are complex social problems and require complex intervention at the societal level, not the individual level.



Conroy, D., Smith, S. and Frethey-Bentham, C. (2018), "Weighing the odds: an exploration of resistance to obesity and overweight", Journal of Social Marketing, Vol. 8 No. 4, pp. 421-441. https://doi.org/10.1108/JSOCM-06-2018-0056

Download as .RIS



Emerald Publishing Limited

Copyright © 2018, Emerald Publishing Limited


In 2018, we have surpassed the population landmark of 7.5 billion people on the panet (Lutz, 2017), and while attention is being paid to the potential of population increase outstripping food supply and thus contributing to an unstable future, more limited attention is being paid to the potential contribution overweight and obesity, a metabolic health crisis, could make to this situation. Overweight and obesity can negatively affect the health of the individual, may restrict positive life experiences and place an increased burden on health systems (van den Hoek et al., 2017). Clearly, there is a need to address the current obesity epidemic that is being reported (Ravussin and Ryan, 2018) to contribute to a more stable future for all nations.

Obesity and being overweight are complex social problems that have developed rapidly over the last 50 or so years (Popkin et al., 2012). Dramatic changes in Western society during this period have undoubtedly contributed to their growth. For example, the food processing technology developed during Second World War has expanded significantly to the point now where processed foods are a regular and daily part of most people’s diets (Poti et al., 2015). As a result, we are increasingly living in an obesogenic environment; i.e. the combined influence of surroundings, opportunities, or life conditions on the promotion of obesity (Boyd and Egger, 2002; Lake and Townshend, 2006). The majority of us are bombarded by cues to eat throughout the day and night and food has become too readily available (Kemps et al., 2014). Moreover, obesity and being overweight are becoming concerns for every nation; for example, in Malaysia the number of obese adults has more than tripled over the past decade (Samy, 2010).

Despite New Zealand, where this study is based, having the third highest adult obesity rate in the OECD, the government has, to date, considered obesity and overweight to be an individual issue and responsibility and not a social issue requiring its intervention (Ng et al., 2014). Yet the problems associated with obesity and overweight are significant social problems requiring a social ecosystem marketing approach (Brennan et al., 2016). Metabolic health concerns such as diabetes and cardiac disease are directly related to weight, and cost the global economy an estimated $US2tn ($NZ 3.1tn) annually, both in health care and the loss of working hours of these individuals (McKinsey and Company, 2014).

The Food and Beverage Industry has suggested it is willing to self-regulate without the need for government intervention, but has so far contributed little (Vandevijvere and Swinburn, 2015). Individuals are looking for answers to the obesity problem but are finding the available advice confusing (Nagler, 2014). There is a need for leadership and for governments to provide advice on how best to proceed to ensure a healthy future for citizens. We propose that the Socio-Ecological Model (SEM), a theoretical framework which recognizes that most public health challenges are too complex to be adequately explored from a single level of analysis, may be useful in analyzing how best to address the obesity epidemic (Stokes, 1996). SEM asserts that change strategies often neglect the social and environmental context in which behaviors occur and are reinforced, suggesting the importance of progams that target multiple levels of intervention (Green et al., 1996).

Research aims

Our research considers the roles of group influence, motivation and self-efficacy to inform new ways of advocating sustained weight loss and maintenance, which will be of relevance to industry, policymakers and consumers. Rather than continuing to focus on the pathology and disease of this phenomenon (overweight, obesity), we propose a focus on the health and resilience of those who resist choices that lead to these conditions. This study considers overweight and obesity from a collective, social perspective, rather than from the individual perspective that currently dominates social marketing campaigns.

By exploring peoples’ engagement with this phenomenon in a social context we aim to better understand which constructs can be used to inform initiatives in the area of healthy weight maintenance. Our objective is to better understand consumer behavior by exploring resilience within the obesogenic environment.

This study is a two-phase mixed methods study. In phase one, a qualitative study, we worked with young adults in a stage of transition, i.e. those who are only just starting to be responsible for all of their nutrient choices and are only just starting to have the financial means to make deliberate choices in how to allocate their spending. The substantive research question for this stage of our research is:


How can a better understanding of resilience to an obesogenic environment assist social marketers to develop successful policy, marketing communications and interventions to achieve a sustainable healthy weight for our nation?

In phase two of the study, the results of the first phase were used to inform the development of a questionnaire, which was administered to a wider and more demographically representative sample of respondents within New Zealand. The substantive research question for this phase of our research is:


What factors enable some, but not all, people to resist the trend for an increasing number of New Zealanders to become overweight or even obese?

Theoretical perspectives and concept development

Models of health promotion have undergone several generational changes. Early fear promotions aimed at dissuading people from performing unhealthy behaviors have been largely discredited (Bandura, 1998). These were replaced with models that focused on extrinsic rewards for healthy behavior, but the behavior was often not sustained once the rewards ceased (Holroyd and Creer, 1986). The next generation of interventions were directed toward the development of self-regulatory capabilities and for many governments, individual responsibility is still the focus (Bandura, 1997). However, contemporary health promotion models acknowledge that personal change occurs within a network of social influences, and such models advocate socially oriented interventions (Bandura, 2004). Contemporary models also note that the importance of the wider environment, for example the neighborhood one lives in, the work one does, the adequacy of the home kitchen, etc. in relation to food intake (Wansink, 2004).

The SEM is widely adopted in health research (Gregson et al., 2001). It provides a framework for describing individual change within the context of social change, and conceptualizes the social world in five levels of influence (Figure 1). These levels are social structure, policy and systems; community; institutional/organizational; interpersonal and individual. This article focusses on the individual and interpersonal levels, noting that a more astute understanding of these levels, and how they interact with social structure, policy and systems, community and institutional/organizational with help to inform beneficial change at these levels.

It is our contention that a focus on increasing self-efficacy and fully recognizing the importance of social groups is the way forward for improving control on overweight, obesity and associated metabolic diseases. This focus has direct links to the SEM framework’s interpersonal and individual community levels. To this end, we argue below that social cognitive theory (SCT) and social identity theory (SIT) are valuable lenses through which to view this research (Holt, 1931; Tajfel and Turner, 1979), and that a better understanding of these levels will result in more informed intervention at the other SEM levels, resulting in improved metabolic health within the community.

Identity, self-concept and efficacy

An individual’s self-concept is multifaceted and these facets change through evaluation, experiences, interaction, etc. (Sirgy, 1982). A person’s identity is the overall perception they have of themselves, and a sense of self is defined by a person’s unique individual characteristics (Hewitt, 1976; Markus and Wurf, 1987). “Dieting” occurs when a person recognizes a negative difference between their perceived and ideal self (Mask and Blanchard, 2011).

SCT posits a multifaceted causal structure in which self-efficacy beliefs operate together with goals, outcome expectations and perceived environmental facilitators in regulating human beliefs and behavior (Bandura, 1994). SCT asserts that belief in one’s efficacy to exercise control is a common pathway through which psychosocial influences affect health functioning. Although commonly considered a model of self-empowerment, SCT also fully acknowledges the importance of social influences, recognizing that in many activities people compare themselves and their performance to that of others, or to standard norms in a society (Bandura, 1994). SCT extends the concept of human agency to collective agency by acknowledging that people do not operate in isolation.

Social identity and reference groups

Reference groups are the people that an individual refers to when evaluating their self (Thompson and Hickey, 2005). Groups offer people a sense of value, belonging, and self-worth (Stets, 2006). According to SIT, we form social identities based on the groups to which we belong, using criteria that defines how all members in our group are similar to one another (in-group) and different to others (out-group) (Tajfel and Turner, 1986). Groups can motivate satisfaction with weight, but also dissatisfaction with weight, and behavioral change towards reducing weight (Strong and Huon, 1999).

The motivations underlying weight loss and maintenance are often both cognitive and social, because of cultural and societal influences; therefore, it is crucial to consider the role of groups in influencing attitudes and behaviors regarding weight, in addition to the role of identity and self-efficacy (Stryker and Burke, 2000). By examining the relationship between role identities and social identities, which often operate simultaneously in a dynamic structural society, we aim to discover links between status, identity, efficacy and lifestyle consumption in the area of overweight and obesity. These insights can then be used to better inform other social systems on how to more effectively approach and intervene in the current metabolic health crisis.

Phase one: qualitative study methodology and methods

In phase one of this study, we adopted the non-dualist theoretical framework of phenomenography developed by Marton (1986). In phenomenographic research, the researcher chooses to study how people experience a given phenomenon, rather than studying the phenomenon itself (Hazel, Conrad, and Martin, 1997). In our study, we explored how participants experience the phenomenon of “weight” and their experience of the impact it has on their identity and status, the identity and status of others, and the influence of self-efficacy on resilience or indulgence. We are not interested in weight loss per se but in how young adults make sense of their resilience in an obesogenic environment.

Bandura (2001) states that too much health knowledge has been accumulated from research studying refractory cases but ignoring successful self-changers. Mindful of this lens, and in keeping with the work of Sara Lawrence-Lightfoot who pioneered portraiture, a qualitative approach that resists social science’s focus on “pathology and disease rather than on health and resistance” (Lawrence-Lightfoot and Davis, 1997, p. 8), we intentionally interviewed only those who were maintaining a healthy weight (indicated by their BMI index) within an obesogenic environment. BMI range provides an indication of resilient behavior. We acknowledge this measure is only one measure that could be considered but it is a standard measure which is still used to reflect adiposity and is especially useful to use within a younger demographic (Riedl et al., 2016). The participants were from a range of backgrounds (working, at home with a child, studying), ethnicities (Pacific Islander, Maori, NZ European, other) and locales within New Zealand.

Given the emergent nature of this work, the intention was to describe and generate a deeper understanding of the different and similar ways individuals make sense of this particular phenomenon (resistance to the obesogenic environment). Our intention in taking this approach of freedom and discovery was to generate several avenues for future exploration in phase two and to facilitate a broad understanding of individuals’ sense-making. Thus, an interpretivist approach was adopted based on in-depth interviews and observations (Smith et al., 2002):

  • Participants: A purposive sampling strategy was used to recruit 31 young people of a normal, healthy weight (Eisenhardt, 1989; Miles and Huberman, 1994; Green, 2005). To address all socio-economic sites, three areas in Auckland, the largest and most diverse city in New Zealand, were identified along with two other regions in New Zealand – Hamilton and the Bay of Plenty. Using a snowballing technique, we recruited both male and female participants between the ages of 17 and 26 (Taylor and Bogdan, 1998).

  • Interviews: Each participant was interviewed in their own home or in their local neighborhood, so that interviewees felt comfortable and could refer to their local surroundings if and when relevant. Interviews lasted approximately 60 min and all interviews were audiotaped and later transcribed. Interviewers kept field note diaries and noted any relevant observations during the interview. At this exploratory stage, interviewers sought to encourage participants to tell stories of their experiences and to express their thoughts about food, health, weight, friends, family, culture, environment and experiences (Åkerlind, 2005). Projective techniques and laddering were used to discover the dominant constructs and values motivating participants’ choice of experiences to share (Reynolds and Gutman, 1988).

  • Analysis: Phenomenography treats all participants as a collective group rather than as independent individuals (Åkerlind et al., 2005) and assumes that while meanings may vary within and between individuals, there are a limited number of qualitatively different understandings of the world (Marton, 1986). In addition to searching for variation in meanings, a phenomenographic analysis also seeks the structural relationships between variations in meanings (Åkerlind, 2012). Thus, to analyze phase one, an iterative approach (Spiggle, 1994) was used to move through the data, viewing individual voices as part of a larger collective narrative.

Because of the nature and goal of phenomenography, the analysis of our research data adopts an “inductive” and “data driven” approach (Boyatzis, 1998). Specifically, inductive thematic analysis was used to identify, code and categorize key themes (i.e. different ways of experiencing resilience) in the data (Boyatzis, 1998). Consequently, identified themes were strongly grounded in the data (Patton, 1990). Furthermore, both open coding and axial coding techniques were employed to strengthen the coding process (Strauss and Corbin, 1998). In keeping with Sandbergh’s (1997) suggestion that interpretive awareness is a worthy way of establishing trustworthiness in phenomenographic research, two of the researchers independently coded the data. This step was taken to maintain an objective awareness by acknowledging and explicitly dealing with each researcher’s subjectivity by contesting and agreeing themes.


The main themes which we identified are as follows: personal factors, understanding of health/physical impacts, media influences, emotional associations, social influences, time/schedule influences, financial resources and environmental factors. These findings, though not unexpected, did reveal an unexpected overarching theme of mindfulness by which we mean a state of active, open attention on the present. When one is mindful, one observes one’s thoughts and feelings from a distance, without judging them as good or bad (Brown and Ryan, 2003). Essentially, one is aware and present. Our participants were not consistently virtuous in their food choices and exercise habits, but they were homogeneously mindful of the consequences that their choices and behaviors had (e.g. “[I know it’s bad to eat late at night] but I don’t have the, the willpower to follow it myself so I still eat late or whenever I feel like it”, Julie).

Personal factors were collectively an important theme and included personal circumstances (e.g. “I did a cooking course […] and I used to work as a chef”, Sarah), personal lifestyle (e.g. “Since I was young I’ve been running marathons and half marathons”, Ben), perceived genetic or physiological predispositions (e.g. “I know that there’s like a high risk of diabetes in my family”, James), personal qualities (e.g. “I have a hard time stopping myself from eating, even though I know I should”, Julie) and beliefs or attitudes (e.g. “I don’t like leaving food behind”, Hugo). These perceived or identified personal factors are strongly linked to self-concept. In turn, self-concept plays a significant role in influencing consumer behavior (Goldsmith et al., 1996). These kinds of perceptions were indicative of the individual stage of the SEM framework, and provided an excellent understanding of the factors which influence behavior. In addition, these factors are seen as crucial to capture within the planned larger study of phase two.

Understanding of health/physical impacts of consumption choices was a second theme. An understanding of how eating impacted respondents physically was expressed in terms of the importance of eating and the impacts of certain food choices (e.g. “if I have unhealthy food it makes my skin worse”, Sarah), the (un)importance of portion size (e.g. “portion size doesn’t matter for me, like, I’ll just eat and eat and eat until I’m full”, Luisa) and the consumption of take away or junk foods (e.g. “I don’t particularly like fizzy drinks […] probably, oh ‘cos I think they’re kind of bad for you so I just don’t drink them”, Sherna). As well as being linked to self-concept, this theme also shows the significant role that self-efficacy plays in determining the belief a person has in their ability to exercise control or at least acknowledge the consequences of losing control (Bandura, 2001). On the whole, the participants in phase one displayed a high level of self-efficacy, but efficacy differs in any population and is most vulnerable in children (Wigfield and Eccles, 2000). We determined, therefore, that level of self-efficacy was an important factor that needed to be examined in phase two, though we acknowledge that having an understanding of health/physical aspects of food choices is only one aspect of self-efficacy. Self-efficacy has significant implications for the SME framework’s “Community” level which speaks to the public agenda, including media agenda. Our data suggest that resilience to mass advertising of nutrient deficient foods and beverages requires high levels of self-efficacy in an obesogenic environment.

Media influences was a third theme. This was expressed as a relationship between self-perception and exposure to media (e.g. “[…] I don’t feel, like, that great about my weight […] It’s mainly to do with […] what I see on TV”, Luisa), learning new skills (e.g. Kathleen learned how to cook in part through the cooking channel), the influence of advertising (e.g. “I would say TV [advertising] would have the most influence [on my food choices]”, Jarasporn) and social media (e.g. “[…] when you’re flicking through Facebook, you know, things that [come] up […] you might not click, but that’s just reinforcing the idea into your mind”, Pente). In addition, food labeling was also viewed as an influential factor (e.g. “[…] the biggest aspect through which advertising influences my buying choices is just, like, [the] esthetic of the product and the way they’ve designed the label”, Hugo). SIT and the influence of reference groups are clearly evident in this theme. For example, while Luisa is using media personalities to identify groups she does not belong to, this is having a negative impact on her sense of self-worth, Kathleen is using media to positively identify with a particular lifestyle – that of a “competent cook”. Self-concept is also influencing the consumption behavior of our participants in terms of their product choices through packaging, and reviews of restaurants by perceived experts. The implication here is that media has a significant and wide-ranging impact on individuals. All of our participants were mindful of this impact and consequently filtered its consequences, but this may not be the case for the wider population. Here the social media influences that are clearly linked to “media influences” have been included under the broader theme of “media”, but the overlap between the social media influences and social influences is acknowledged. In addition to this theme reflecting the SEM level of “Community” (under which media resides), there are also implications for the levels of “Institutional”, for example, the role of regulation in food labeling and social structure, e.g. the role of policy in food regulation.

Emotional associations with food consumption were also apparent from the data. Positive and negative emotions were associated with eating (e.g. “[…] it’s relaxing, soothing, it’s like something that you do when you’re bored”, Brooke; “Like food was bound up with anxiety and stress about what I was eating”, Giles), cooking (e.g. “[…] when I’m cooking it’s [kind of] like in my own sort of space. It’s kind of room to experiment”, Nick; “[…] if you don’t have a particular ingredient you have to improvise and I can’t do that and so I don’t and so I get stressed out”, Sherna) and shopping (“I think [food shopping is] just, like, a stress killer sometimes”, Jarasporn). Our participants were mindful and aware of the pressure and pleasures associated with food, but again, this may not necessarily be the case in the wider community.

Social influences included such themes as upbringing or family norms (e.g. “I think it’s the food that we eat [that causes weight issues in my family]” Luisa), socializing and food consumption (e.g. “That’s kind of how we socialise and catch up is to go out for coffees and go out for dinner and things like that”, Tasmin), the influence of others on food choices (e.g. “I was following all these vegans on Instagram and I was, there was quite a big, like, vegan following. And they have a little community so I was trying out different recipes through that”, Julie), the impact of others’ opinions (e.g. “then she, like, met her first boyfriend […] he was also obsessed with the way he looked so she was also obsessed with the way she looked”, Sherna) and consideration of others (e.g. “I also want my partner to stay healthy so often I will want, I might lean towards getting a take-away or something, but you know, he may have been, you know, for a while he was not on a diet […] and so I would not get takeaways, to be encouraging”, Jessica). The motivations underlying weight loss and maintenance are often both cognitive and social, because of cultural and societal influences (Stryker and Burke, 2000); therefore, we argue that it is crucial to consider the role of groups in influencing attitudes and behaviors regarding weight, in addition to the role of identity and self-efficacy. We determined, therefore, that the role social influences play in the wider population’s consumption choices needs to be explored in relation to resilience within an obesogenic environment. This finding is in keeping with the “Interpersonal” level of the SEM framework which stresses that individuals are embedded within a network of primary groups, e.g. family and social networks, which help shape social identity and assist with personal role definitions.

Organizational skills was another theme, and encapsulated sub-themes such as time restraints (e.g. “some days I feel like eating’s an inconvenience, and it takes up too much time […]”, Giles), which invariably lead to some positive behaviors (e.g. “I think [we started using meal plans] because my step-mum started studying and so they didn’t want to have to spend so much time thinking, like, planning their meals for the next, like for the week”, Daisy) and some negative (e.g. “like if I’m late for work something […] like I’ll just pick something up from McDonalds or something like that”, Ramari). Again, the role of self-efficacy is clearly evident in that our participants were largely mindful of the need to be organized. Citizens with lower self-efficacy are more likely to feel overwhelmed if they do not have some intervention that raises their empowerment level (Ashford and LeCroy, 2009). Again, this theme has implications for the SEM framework’s levels of “Institutional” which focuses, for example, on informal structures such as schools which could push an agenda to empower people’s skills; the “Community” which, for example, could demand social change, and “Policy” which could, for example regulate or support healthy actions such as providing budgeting skills in schools.

Financial resources were also an important factor for the respondents in that consumption habits were related to their abundance or lack thereof. Most (if not all) respondents discussed the way financial restraints contributed to food choices (e.g. “If I go to a cafe I’ll always pick something sweet, it’s just [because], one, it’s cheaper; because if you pick a savory thing in a cafe it’s always [going to] be more expensive than a sweet thing”, Luisa); the relationship between enjoying food and perceived cost (e.g. “[Food shopping would be more enjoyable] if prices dropped down a bit on some things”, Pente); and the link between cost and food quality (e.g. “It’s really difficult to find something healthy around the university that’s not, like, ridiculously expensive”, Daisy). Our participants were young adults and so tended to be on a restricted budget; thus it is not unexpected that they were aware of the costs of food. However, despite previous work suggesting that young adults are susceptible to the environment (Laska et al., 2010), we found our participants exhibited higher than anticipated involvement with shopping and food choices, again demonstrating high levels of efficacy and mindfulness. As with the previous theme, the links to the SEM levels of “Social Structure”, “Community” and “Institutional” are clear if empowerment of citizens is to be attained.

Environmental factors, another emergent theme, is clustered around the household environment (e.g. “Our oven and stove are really messed up at this house and we’re getting the kitchen renovated soon, but right now it’s really difficult to cook”, Hugo); the local neighborhood (e.g. “I used to be a member at a gym, but then that gym got closed down and there wasn’t any gym nearby” Ramari); work/study environments (e.g. “[I usually have portioned food in] study environments, [because] you’re kind of stress eating”, Pente); and other situational factors (e.g. “[…] and I would take all the confectionaries away from the front [at the supermarket]”, Atah). These sub-themes seem to be contingent factors that affect (e.g. stimulate or inhibit) behavior. Certainly, the inclusion of environmental factors is an acknowledgement of the important role they play, for example in relation to food intake (Wansink, 2004), alongside, and integrated with, individual and social factors. Further, there is also evidence for intervention that allows for places of work/study to supply nutritious reasonably priced foods (level of “Community”), which may require a change in regulation (level of “Institutional”), which may only be possible by a change in policy (level of “Social Structure”).

These themes and sub-themes are summarized in Table I. The table also links these findings with the survey elements developed for the quantitative phase of this project. The survey items were developed through a reading of the literature in addition to the analysis of the phase one data.

Phase two: quantitative survey methodology and methods

Findings from the qualitative phase were utilized to develop a quantitative questionnaire to examine the factors which enable some people to resist the trend toward becoming overweight or obese. The quantitative phase of the study was extended to the broader New Zealand population to better understand the resistance strategies (or lack thereof) adopted by adults 17 years and older, and to allow comparison between those individuals with a healthy BMI (18.5 < 25), and those who were overweight or obese (BMI ≥25). We acknowledge that BMI is not a perfect measure of resilience to the obesogenic environment, as it is subject to a number of influences other than resistance strategies (e.g. genetics) and does not account for muscle mass. However, BMI has been demonstrated to relate to lifestyle choices in a number of prior studies (Cugnetto et al., 2007; Epstein et al., 2008; Green et al., 2015; Sundquist and Johansson, 1998). Furthermore, we are interested in how our participants experience the phenomenon of “weight”, particularly how this experience may impact on their sense of identity and status, and the influence of self-efficacy on resilience or indulgence. Consequently, we also validate the findings by investigating the impact of resilience strategies on BMI with other outcome measures (i.e. “resistance to overeating”, “extent of overweight family members” and “extent of overweight friends”).

Instrument and sample

Drawing on an extensive review of relevant literature and the phase one interviews, an initial sample of n = 34 items to measure resilience was developed. Four expert judges evaluated the scale for conceptual consistency of items, face validity and content validity. The items were subjected to cognitive pretesting on a sample of 15 individuals to evaluate comprehension, retrieval of information, judgement and ability to formulate a response (Tourangeau, 1984; Collins, 2003). To ensure applicability of the survey to the wider New Zealand population, the pretest participants were selected from a wide range of age groups and with varying BMIs. To avoid any potential question order effects, all survey items were randomized in the online administration of the survey.

A sample of n = 959 individuals was recruited to take part in an online survey. As the study objective was to explore differences between overweight or obese adults and those of a healthy weight, those individuals with an unhealthy low BMI (<18.5) were removed from the study (n = 38) leaving a total usable sample of n = 921 individuals, ages 17 years and older. Survey quotas and data weighting were applied with respect to gender, age, ethnicity, location and respondents’ BMI (calculated as bodyweight in kg/height in meters squared) to ensure alignment between the sample and New Zealand population statistics.

Exploratory and confirmatory factor analysis

The 34-scale items were subjected to a principal components exploratory factor analysis (EFA) with orthogonal rotation (Varimax). EFA was performed prior to confirmatory factor analysis (CFA) to identify the underlying factor structure without imposing any preconceived structure on the outcome (Child, 1990). The exploratory factor structure explained 62 per cent of the variance in the data and revealed the existence of nine distinct factors (Table II). Two items with factor loadings <0.4 were excluded from the analysis to enable a clearer interpretation of the factors (Tabachnick et al., 2001; Hair et al., 2014).

CFA using maximum likelihood estimation was subsequently conducted to assess scale dimensionality (Anderson and Gerbing, 1988). In the initial model development phase, four variables demonstrating standardized regression weights less than 0.5 were removed from the model. Less than 20 per cent of items were removed from the CFA model (as recommended by Hair et al. (2014)).

The chi-square test for goodness of fit for the model was significant at the 5 per cent level, χ2 (330) = 1,242.681, p < 0.00; this is not unexpected with a large sample. Findings indicate an acceptable fit for the nine factor structure (GFI = 0.903, CFI = 0.902, RMSEA = 0.056) given the large sample size and larger number of observed variables (Hair et al., 2014). Discriminant validity was established by comparing the AVE with the squared correlations between the factors. Composite reliability estimates are all above 0.6 (with eight of the nine being above 0.7), and all remaining items have significant loadings on their corresponding factor, with standardized regression weights all above 0.5 and most above 0.7 as recommended by Hair et al. (2014). The factors represent inherent attitudes and behaviors that are likely to determine how resilience (or lack thereof) to the obesogenic environment will be manifested (Mehrabian and Russell, 1974).

Cluster analysis

Cluster analysis was used to investigate the resilience strategies of subgroups within the data. This technique has been effective at describing subgroups on similar topics (Green et al., 2015 used cluster analysis to investigate the characteristics of obese individuals). Factor scores were used to generate a cluster analysis. To produce the cluster analysis, data were randomly divided into two subsets (Lockshin et al., 1997; Michaelidou, 2012). The first subset was used to generate the possible alternative cluster solutions using hierarchical cluster analysis with Ward’s method (squared Euclidean distance). The dendrogram and agglomeration coefficients from the hierarchical cluster analysis suggested the usage of a four, five or six cluster solution. The second subset of data was then used to conduct a k-means cluster analysis using four, five and six cluster solutions. The initial centroids provided by the hierarchical analysis were used as cluster seeds in the k-means analysis (Punj and Stewart, 1983; Hair et al., 2014). A hierarchical cluster analysis was also performed on the second subset of data (using Ward’s method, squared Euclidean distance) and compared to the results of the k-means analyses from the same data file. This was conducted to provide an indication of the stability of the solution (Punj and Stewart, 1983). The four-cluster solution was chosen as the most appropriate in terms of stability and reproducibility. The data sets were then combined and a final k-means cluster analysis was conducted. The final cluster solution is presented in Table III.

In comparison to other clusters, Cluster 1 (Strivers; 23 per cent) has higher mean scores for weight focus and routine, and lower mean scores for inability to read body signals, family and cultural importance of food, cooking and preparing meals, and susceptibility to media and unhealthy social influences. Cluster 2 (Susceptible; 21 per cent) has higher mean scores for nearly all factors, with the exception of routine. Cluster 3 (Disengaged; 30 per cent) has higher mean scores on financial factors, susceptibility to media and unhealthy social influences, and inability to read body signals; and lower mean scores on planning and preparation, routine and weight focus. Cluster 4 (Mindful; 26 per cent) has higher mean scores on planning and preparation, routine and cooking and preparing meals and lower scores on weight focus, inability to read body signals and susceptibility to unhealthy social and media influences.

To establish the external validity of the cluster solution, criterion validity was assessed using variables omitted when developing clusters (Hair et al., 2014). BMI, resistance to foods when emotional and reported extent of overweight peers (i.e. both overweight close friends and overweight family members) were chosen to validate the cluster solution because BMI is only one marker of resilience within an obesogenic environment. Emotional eating and weight of peers have also been associated with (lack of) resistance to the obesogenic environment (Neumark-Sztainer et al., 2007; Trogdon et al., 2008; Hruschka et al., 2011).

A MANOVA model was estimated using these four items as dependent variables. The MANOVA model is significant at the 5 per cent level (F = 2421.1, p = 0.00), providing support that these variables can be predicted by cluster membership. Figure 2 displays the sample means from the MANOVA analysis for each cluster. Tukey honestly significant difference (HSD) post hoc tests for differences between groups reveal that the Mindful cluster has a significantly lower BMI and is less likely to report having many overweight family members and close friends in comparison to the other clusters (p = 0.000 for all comparisons). The Strivers group has a significantly lower BMI than that of the Susceptible at the 10 per cent level (p = 0.056). The Mindful and Strivers groups also report significantly higher resistance to emotional eating than the Susceptible (p = 0.000 for both comparisons) and Disengaged (p = 0.000 for both comparisons). The Susceptible group are more likely to report having many overweight family members and close friends in comparison to Strivers (family members, p = 0.001; friends, p = 0.003) and the Disengaged (family members, p = 0.006; friends, p = 0.000).

For descriptive purposes, the clusters were further profiled on a set of additional variables not included in the clustering variate, or used to assess predictive validity. The Strivers are mostly male, mostly of age 45 years and older, mostly of New Zealand European ethnicity and have a relatively even spread in terms of personal income and education levels. The Susceptible group is mostly female, mostly aged below 45 years, comprises a higher proportion of Asians in comparison to the other clusters, and has higher income and education levels than the other clusters. The Disengaged group has an approximately even proportion of males versus females, mostly of age below 45 years and comprises the highest proportion of New Zealand Māori and Pacific Island members. The Disengaged are the group least likely to possess a bachelor’s degree or higher, and have lower income levels than the Strivers and Susceptible. The Mindful group has an approximately even proportion of males versus females and is mostly of age 30 years and older. They have lower income levels than the Strivers and Susceptible, but are the second most highly educated group.

In terms of eating habits, the Mindful cluster are the least likely to have consumed meals prepared outside of the home within the last week, the most likely to dine at home (with other members of their household) regularly, and the most likely to be the main decision maker on meals eaten in their household. Conversely, the Disengaged and Susceptible groups demonstrate a higher incidence of consuming meals prepared outside of the home within the past week, and are least likely to dine with other members of their household. The Disengaged are also the least likely to be the main decision maker on meals eaten in their household.

The Mindful and Strivers groups demonstrate the greatest resistance to unhealthy social influences. In comparison to unhealthy social influences, healthy social influences on eating habits are more commonly reported by all groups. However, the Mindful group are more susceptible to healthy social influences from peers (in particular those they live with), compared to the other groups. The Susceptible group is sensitive to all healthy social influences, in particular advice on the internet and advice from radio, TV, books or magazines. Conversely, the Disengaged are less susceptible to healthy influences in comparison to the other clusters. Tukey HSD post hoc tests also reveal that the Mindful group are more likely to report enjoying dining with others in comparison to the other groups (p = 0.000 for all three comparisons).


The main themes identified in phase one were largely echoed in the phase two factor analysis: weight focus, inability to read body signals, cooking and preparing meals, planning and preparation, routine, unhealthy social influence, financial, media influences and family and culture. This result demonstrates that these factors are concerns for the entire population and supports our contention that overweight and obesity are a social concern and have an impact throughout society; therefore, socially oriented interventions (Bandura, 2004) are more likely to be effective than a focus on the individual. Importantly, the need to consider metabolic disease using the lens of the SME framework was further underlined. The embeddedness of the individual and the interpersonal within wider society, institutional norms and the influence of policy and social structure are clear.

The cluster analysis reveals four distinct groups – Strivers, Susceptible, Disengaged and Mindful – with the Mindful group being most similar in attitudes and behaviors to the resistant young adults interviewed in phase one. Self-efficacy in both the Mindful and resistant participants is very high; they reported feeling in control, organized, educated and empowered by their social networks. This finding resonates with SCT’s assertion that belief in one’s efficacy to exercise control is a common pathway through which psychosocial influences affect health functioning (Schwarzer and Fuchs, 1996). Here, it must be recalled that whilst self-efficacy is often considered to be concerned with the individual, it is understood that its development is governed by the social environment and social contacts of the individual (Bandura, 1994). In contrast, the high levels of efficacy that are required for change to be attempted and sustainable are not in evidence in the other three groups, especially the Disengaged group, which contains the majority of our New Zealand Māori and Pacific Island participants, who come from a more collectivist culture where self-efficacy is more likely to be judged on inter-group behavior rather than individual achievement (Hofstede, 1980). A greater understanding of how to increase the self-efficacy of this group is necessary if any intervention, social or otherwise, is to be successful. However, clearly the Mindful and Strivers groups feel empowered within our society, whilst the disengaged group lacks such empowerment. The wider social structures are not facilitating their health focus and clearly intervention is needed to empower behavioral change.

Research clearly demonstrates that the influence of others can have a significant impact on people’s behavior (Bandura, 2004); the interpersonal level of the SEM framework. Both the resilient young adults and the Mindful group find positive support in their reference groups. Groups offer people a sense of value, belonging and self-worth (Stets, 2006) and lifestyles are a form of identification that people use to both differentiate themselves from and connect to others in their community (Evans and Jackson, 2007). The Mindful group also readily identifies with positive influence in their environment (“Community” level of the SEM framework), and in the media (“Community” level of the SEM framework). It would seem that their ‘in-group’ is anything positively health-related, whilst the low self-efficacy of the Susceptible and Strivers makes them more vulnerable to identifying with less healthy groups. The Disengaged group are less influenced by healthy media and social sources than any other group, but highly influenced by unhealthy social and media sources. This finding suggests that they are highly identified with their own “in-group” social identity, and culture. Clearly further research is necessary to explore why these groups identify with different aspects of media and how other levels of the SEM framework can be best utilized to empower citizens to strive for a healthy metabolic state.


The aim of this study was to better determine a pathway for successful intervention that will improve the health of our citizens. Our concern is also one of providing solutions which will benefit individual health, social health, and health (economic) systems. Our findings suggest that social structures (SEM framework “Interpersonal” level) can positively affect individuals and their sense of self when trying to resist weight gain. Belonging increases feelings of status, self-efficacy and positive identity, suggesting that rather than focusing on the detrimental impact overweight has on the individual’s identity and status, more positive and social approaches to weight loss may have a greater influence on our battle with obesity and desire for a sustainable future. Empowering citizens and increasing efficacy and mindfulness are potential paths forward. Such pathways require a more holistic approach to be taken than simply focusing on the individual; we suggest that the SEM framework is an excellent way to approach this.

The field of health in general and obesity in particular is overwhelmed by contention. Political battles between individualist and structuralist approaches are all too common. Proponents of the individualist agenda claim that individuals can exercise a good deal of control over their health; therefore, it is their own responsibility to maintain it. Those of a more structuralist persuasion argue that health is largely the product of social, environmental, political and economic conditions, and that the individual has very little control over these. Our findings suggests that leadership in the area of obesity needs both approaches if we are to achieve sustained health for all citizens. It is our contention that the quality of health of a nation is a social matter, not just a personal one. Consequently, it requires changing the practices of social systems that impair health, and not simply changing individual habits. The main focus of a social approach to obesity is a collective enablement for changing social, political and environmental conditions that affect overweight and obesity, all found in the SEM framework.

Our main contribution to the conversation addressing increasing levels of overweight and obesity is research findings that demonstrate these are complex social problems and require complex intervention at the societal level, not the individual level. Interventions targeting the individual have failed repeatedly, costing taxpayers millions of dollars, and individual citizens, years of stress, and sometimes their lives. A third phase to this research would be to test our theory by creating an intervention to be experimentally tested. For this phase, we are particularly interested in the Strivers group who are, in many ways, similar to the Mindful group, yet fail in their endeavors, and the Susceptible who are currently vulnerable to negative social messages regarding healthy eating. Additionally, an interesting avenue for future research is the Disengaged group, which would require a focus on the issue of independent versus interdependent societies to better understand how high self-efficacy is socialized, understood and empowered in each society to target interventions more appropriately [Fifita et al.’s (2015 study on resistance to tobacco consumption]. The SEM framework to be of value as a framework for this work.

In conclusion, we urge research and intervention to shift from a lens on the individual to taking a more collective, societal focus that requires significant leadership from government. We assert, as do Kilbourne et al. (1997, p. 20), that the mainstream marketing academy needs to increasingly take on board the issues being discussed by social marketers and others outside our academy. Metabolic health is a highly complex issue and no one discipline has the answers; it is only by working together that complex answers have any hope of emerging.



Figure 1.


Cluster validation sample means

Figure 2.

Cluster validation sample means

Themes associated with resilient group and associated survey items

Themes Sub-themes Survey items
Personal factors Personal circumstances
Perceived predispositions
Personal qualities
Beliefs or attitudes
I find it difficult to stick to diets and/or healthy eating
I over-eat before realizing that I am full
I find it difficult to stop eating, even when I have probably had enough
I think about my weight when choosing what I eat
I set weight loss goals for myself
I think about my weight
When I set weight loss goals, I achieve them
Understanding of health/physical impacts Importance of eating certain foods
(Un)importance of portion size
Consumption of take-away or junk foods
I am aware of how many calories are in the foods that I eat
I monitor how much I eat
I weigh myself and/or take my body measurements
I know how to cook and prepare meals
I diet and/or participate in weight loss programs
I like to try the latest weight loss programs and/or diets
Media influences Self-perception and exposure to media
Learning new skills
Influence of advertising
Social media
Food labelling
I get hungry when I see food advertised
I try new foods I see advertised
I get ideas on what to cook from television shows, the internet, social media and/or advertising
I read food labels before choosing which foods to buy
Emotional associations Positive/negative emotions associated with eating
Positive/negative emotions associated with cooking
Positive/negative emotions associated with shopping
I generally enjoy cooking and preparing meals
When I was a child, I was rewarded with food
Social influences Upbringing/family norms
Socializing and food consumption
the influence of others on food choices
The impact of others’ opinions
Consideration of others
When I am around others, I make unhealthier food choices than I would if I was alone
I attend functions, gatherings or celebrations where there is unhealthy food
Food has always been important in my family
Food is central to my culture
I enjoy cooking for others
Organizational skills Time-restraints
Meal planning
Consumption of convenience food
I plan my meals in advance
I take time to prepare healthy meals
I keep healthy snacks on hand for when I get hungry
I skip meals
I eat breakfast
I generally eat meals at the same time each day
Financial resources Resources affect food choices
Resources affect enjoyment of food/shopping
Link between cost and food quality
I can’t afford to eat healthy food
Healthy food is expensive
Environmental factors Household environment
Local neighborhood
Work/study environments
Other situational factors
Healthy food is readily available to me (e.g. at home, at my place of work, etc.)

Rotated component matrix

Survey items Component
1 2 3 4 5 6 7 8 9
I think about my weight 0.623                
I am aware of how many calories are in the foods that I eat 0.558                
I weigh myself and/or take my body measurements 0.734                
I monitor how much I eat 0.645                
I think about my weight when choosing what I eat 0.676                
I set weight loss goals for myself 0.805                
I diet and/or participate in weight loss programs 0.620                
I like to try the latest weight loss programs and/or diets 0.421                
When I set weight loss goals, I achieve them (reversed)   0.435              
I find it difficult to stick to diets and/or healthy eating   0.565              
I over-eat before realizing that I am full   0.816              
I find it difficult to stop eating, even when I have probably had enough   0.801              
I generally enjoy cooking and preparing meals     0.860            
I know how to cook and prepare meals     0.707            
I enjoy cooking for others     0.856            
I plan my meals in advance       0.633          
I take time to prepare healthy meals       0.627          
I keep healthy snacks on hand for when I get hungry       0.676          
I read food labels before choosing which foods to buy       0.641          
I eat breakfast         0.762        
I skip meals (reversed)         0.800        
I generally eat meals at the same time each day         0.699        
My eating habits are influenced by other people           0.406      
When I am around others, I make unhealthier food choices than I would if I was alone           0.538      
I attend functions, gatherings or celebrations where there is unhealthy food           0.764      
I can’t afford to eat healthy food             0.784    
Healthy food is expensive             0.858    
I get hungry when I see food advertised               0.582  
I try new foods I see advertised               0.836  
I get ideas on what to cook from television shows, the internet, social media and/or advertising               0.551  
Food has always been important in my family                 0.820
Food is central to my culture                 0.771

Cluster solution

Themes Clustersa ANOVA (F) p
1 2 3 4
Weight focus 3.1 3.4 2.1 2.3 111.896 0.000
Inability to Read Body signals 2.4 3.2 3.2 2.1 88.322 0.000
Media influences 2.3 3.4 2.8 2.5 59.487 0.000
Cooking and preparing meals 3.3 3.9 3.5 4.2 52.781 0.000
Routine 3.5 3.2 2.8 3.8 49.844 0.000
Planning and preparation 3.3 3.4 2.4 3.5 49.844 0.000
Unhealthy social influence 2.5 3.3 2.8 2.4 48.892 0.000
Financial 3.0 3.5 3.6 2.7 40.649 0.000
Family and culture 3.0 3.7 3.3 3.5 23.813 0.000

Cluster means are based on overall scores. Scores range from 1-5 (where 1 = low and 5 = high)


Åkerlind, G. (2005), “Learning about phenomenography: interviewing, data analysis and the qualitative research paradigm”, in Bowden, J.A. and Green, P. (Eds), Doing Developmental Phenomenography, RMIT University Press, Melbourne, pp. 63-73.

Åkerlind, G. (2012), “Variation and commonality in phenomenographic research methods”, Higher Education Research and Development, Vol. 31 No. 1, pp. 115-127.

Åkerlind, G., Bowden, J.A. and Green, P. (2005), “Learning to do phenomenography: a reflective discussion”, in Bowden, J.A. and Green, P. (Eds), Doing Developmental Phenomenography, RMIT University Press, Melbourne, pp. 74-100.

Anderson, J.C. and Gerbing, D.W. (1988), “Structural equation modeling in practice: a review and recommended two-step approach”, Psychological Bulletin, Vol. 103 No. 3, pp. 411-423.

Ashford, J.B. and LeCroy, C.W. (2009), Human Behavior in the Social Environment: A Multidimensional Perspective, 4th ed., Wadsworth Cengage Learning, Belmont, CA.

Bandura, A. (1994), “Self-efficacy”, inRamachaudran, V.S. (Ed.), Encyclopedia of Human Behaviour, Academic Press, London, pp. 71-81.

Bandura, A. (1997), Self-Efficacy: The Exercise of Control, Freeman, New York, NY.

Bandura, A. (1998), “Health promotion from the perspective of social cognitive theory”, Psychology and Health, Vol. 13 No. 4, pp. 623-649.

Bandura, A. (2001), “Social cognitive theory: an agentic perspective”, Annual Review of Psychology, Vol. 52 No. 1, pp. 1-26.

Bandura, A. (2004), “Health promotion by social cognitive means”, Health Education & Behavior: The Official Publication of the Society for Public Health Education, Vol. 31 No. 2, pp. 143-164.

Boyatzis, R.E. (1998), Transforming Qualitative Information: Thematic Analysis and Code Development, Sage, CA.

Boyd, S. and Egger, G. (2002), “Preventive strategies against weight gain and obesity”, Obesity Reviews, Vol. 3 No. 4, pp. 289-301.

Brennan, L., Previte, J. and Fry, M.L. (2016), “Social marketing’s consumer myopia: applying a behavioural ecological model to address wicked problems”, Journal of Social Marketing, Vol. 6 No. 3, pp. 219-239.

Brown, K.W. and Ryan, R.M. (2003), “The benefits of being present: mindfulness and its role in psychological well-being”, Journal of Personality and Social Psychology, Vol. 84 No. 4, pp. 822-848.

Child, D. (1990), The Essentials of Factor Analysis, 2nd ed., Cassel Educational Limited, London.

Collins, D. (2003), “Pretesting survey instruments: an overview of cognitive methods”, Quality of Life Research: an International Journal of Quality of Life Aspects of Treatment, Care and Rehabilitation, Vol. 12 No. 3, pp. 229-238.

Cugnetto, M.L., Saab, P.G., Llabre, M.M., Goldberg, R., McCalla, J.R. and Schneiderman, N. (2007), “Lifestyle factors, body mass index, and lipid profile in adolescents”, Journal of Pediatric Psychology, Vol. 33 No. 7, pp. 761-771.

Green, M.A., Strong, M., Razak, F., Subramanian, S.V., Relton, C. and Bissell, P. (2015), “Who are the obese? A cluster analysis exploring subgroups of the obese”, Journal of Public Health, Vol. 38 No. 2, pp. 258-264.

Eisenhardt, K.M. (1989), “Building theories from case study research”, Academy of Management Review, Vol. 14 No. 4, pp. 532-550.

Epstein, L.H., Roemmich, J.N., Robinson, J.L., Paluch, R.A., Winiewicz, D., Fuerch, J. and Robinson, T. (2008), “A randomized trial of the effects of reducing television viewing and computer use on body mass index in young children”, Archives of Pediatrics and Adolescent Medicine, Vol. 162 No. 3, pp. 239-245.

Evans, D. and Jackson, T.D. (2007), “Towards a sociology of sustainable lifestyles”, RESOLVE Working Paper Series, Vol. 3.

Fifita, I.M.E., Smith, S.D. and Fernandez, K.V. (2015), “The role of self-construal in resisting tobacco smoking”, Australasian Marketing Journal, Vol. 23 No. 4, pp. 294-302.

Goldsmith, R.E., Flynn, L.R. and Moore, M.A. (1996), “The self-concept of fashion leaders”, Clothing and Textiles Research Journal, Vol. 14 No. 4, pp. 242-248.

Green, P. (2005), “A rigorous journey into phenomenography: from a naturalistic inquirer standpoint”, in Bowden, J.A. and Green, P. (Eds), Doing Developmental Phenomenography, RMIT University Press, Melbourne, pp. 32-46.

Green, L., Richard, L. and Potvin, L. (1996), “Ecological foundations for health promotion”, American Journal of Health Promotion, Vol. 10 No. 4, pp. 270-281.

Gregson, J., Foerster, S.B., Orr, R., Jones, L., Benedict, J., Clarke, B., Hersey, J., Lewis, J. and Zotz, K. (2001), “System, environmental, and policy changes: using the social-ecological model as a framework for evaluating nutrition education and social marketing programs with low-income audiences”, Journal of Nutrition Education, Vol. 33, pp. 4-15.

Hair, J.F., Black, W.C., Babin, B.J. and Anderson, R.E. (2014), Multivariate Data Analysis: Pearson New International Edition, 7th ed., Pearson Education, Upper Saddle River, NJ.

Hazel, E., Conrad, L. and Martin, E. (1997), “Exploring gender and phenomenography”, Higher Education Research and Development, Vol. 16 No. 2, pp. 213-266.

Hewitt, J.P. (1976), Self and Society: A Symbolic Interactionist Social Psychology, Allyn and Bacon, Boston, MA.

Hofstede, G. (1980), Culture’s Consequences: International Differences in Work-Related Values, Sage Publications, Beverly Hills, CA.

Holroyd, K.A. and Creer, T.L. (1986), Self-Management of Chronic Disease: Handbook of Clinical Interventions and Research, Academic Press, London.

Holt, E.B. (1931), Animal Drive and the Learning Process: An Essay toward Radical Empiricism, Henry Holt and Co, New York, NY.

Hruschka, D.J., Brewis, A.A., Wutich, A. and Morin, B. (2011), “Shared norms and their explanation for the social clustering of obesity”, American Journal of Public Health, Vol. 101 No. S1, pp. 295-300.

Kemps, E., Tiggemann, M. and Hollitt, S. (2014), “Exposure to television food advertising primes food-related cognitions and triggers motivation to eat”, Psychology & Health, Vol. 29 No. 10, pp. 1192-1205.

Kilbourne, W., McDonagh, P. and Prothero, A. (1997), “Sustainable consumption and the quality of life: a macromarketing challenge to the dominant social paradigm”, Journal of Macromarketing, Vol. 17 No. 1, pp. 4-24.

Lake, A. and Townshend, T. (2006), “Obesogenic environments: exploring the built and food environments”, The Journal of the Royal Society for the Promotion of Health, Vol. 126 No. 6, pp. 262-267.

Laska, M.N., Graham, D.J., Moe, S.G. and Van Riper, D. (2010), “Young adult eating and food-purchasing patterns: food store location and residential proximity”, American Journal of Preventive Medicine, Vol. 39 No. 5, pp. 464-467.

Lawrence-Lightfoot, S. and Davis, J.H. (1997), The Art and Science of Portraiture, Jossey Bass, San Francisco, CA.

Lockshin, L.S., Spawton, A.L. and Macintosh, G. (1997), “Using product, brand and purchasing involvement for retail segmentation”, Journal of Retailing and Consumer Services, Vol. 4 No. 3, pp. 171-183.

Lutz, W. (2017), “How population growth relates to climate change”, Proceedings of the National Academy of Sciences of the United States of America, Vol. 114 No. 46, pp. 12103-12105.

McKinsey and Company (2014), “Overcoming obesity: an initial economic analysis”, Discussion paper, available at: file:///C:/Users/dcon010/Downloads/MGI_Overcoming_obesity_Full_report.pdf

McLeory, K.R., Bibeau, D., Steckler, A. and Glanz, K. (1988), “An ecological perspective on health promotion programmes”, Health Education Quarterly, Vol. 15 No. 4, pp. 351-377.

Markus, H. and Wurf, E. (1987), “The dynamic self-concept: a social psychological perspective”, Annual Review of Psychology, Vol. 38 No. 1, pp. 299-337.

Marton, F. (1986), “Phenomenography: a research approach investigating different understandings of reality”, Journal of Thought, Vol. 21 No. 3, pp. 28-49.

Mask, L. and Blanchard, C.M. (2011), “The protective role of general self-determination against ‘thin ideal’ media exposure on women’s body image and eating-related concerns”, Journal of Health Psychology, Vol. 16 No. 3, pp. 489-499.

Mehrabian, A. and Russell, J.A. (1974), An Approach to Environmental Psychology, MIT Press, Cambridge, MA.

Michaelidou, N. (2012), “A typology of consumers’ variety-seeking disposition based on inherent needs”, Journal of Marketing Management, Vol. 28 Nos 5/6, pp. 676-694.

Miles, M.B. and Huberman, A.M. (1994), Qualitative Data Analysis: An Expanded Source Book, SAGE Publications, Thousand Oaks, CA.

Nagler, R.H. (2014), “Adverse outcomes associated with media exposure to contradictory nutrition messages”, Journal of Health Communication, Vol. 19 No. 1, pp. 24-40.

Neumark-Sztainer, D.R., Wall, M.M., Haines, J.I., Story, M.T., Sherwood, N.E. and van den Berg, P.A. (2007), “Shared risk and protective factors for overweight and disordered eating in adolescents”, American Journal of Preventive Medicine, Vol. 33 No. 5, pp. 359-369.

Ng, M., Fleming, T., Robinson, M., Thomson, B., Graetz, N., Margono, C., Mullany, E.C. et al. (2014), “Global, regional, and national prevalence of overweight and obesity in children and adults during 1980-2013: a systematic analysis for the global burden of disease study 2013”, The Lancet, Vol. 384 No. 9945, pp. 766-781 .

Patton, M.Q. (1990), Qualitative Evaluation and Research Methods, SAGE Publications, Thousand Oaks, CA.

Popkin, B.M., Adair, L.S. and Ng, S.W. (2012), “Global nutrition transition and the pandemic of obesity in developing countries”, Nutrition Reviews, Vol. 70 No. 1, pp. 3-21.

Poti, J.M., Mendez, M.A., Ng, S.W. and Popkin, B.M. (2015), “Is the degree of food processing and convenience linked with the nutritional quality of foods purchased by US households?”, The American Journal of Clinical Nutrition, Vol. 101 No. 6, pp. 1251-1262.

Punj, G. and Stewart, D.W. (1983), “Cluster analysis in marketing research: review and suggestions for application”, Journal of Marketing Research, Vol. 20 No. 2, pp. 134-148.

Ravussin, E. and Ryan, D.H. (2018), “Three new perspectives on the perfect storm: what’s behind the obesity epidemic?”, Obesity (Silver Spring, Md.), Vol. 26 No. 1, pp. 9-10.

Riedl, A., Vogt, S., Holle, R., de las Heras Gala, T., Laxy, M., Peters, A. and Thorand, B. (2016), “Comparison of different measures of obesity in their association with health-related quality of life in older adults–results from the KORA-Age study”, Public Health Nutrition, Vol. 19 No. 18, pp. 3276-3286.

Reynolds, T.J. and Gutman, J. (1988), “Laddering theory, method, analysis, and interpretation”, Journal of Advertising Research, Vol. 28 No. 1, pp. 11-31.

Samy, F.A. (2010), “Malaysians getting obese by eating too heavily at night”, available at: www.thestar.com.my/news/nation/2010/04/11/malaysians-getting-obese–by-eating-too-heavily-at-night/ (accessed 27 February 2016).

Sandbergh, J. (1997), “Are phenomenographic results reliable?”, Higher Education Research and Development, Vol. 16 No. 2, pp. 203-212.

Schwarzer, R. and Fuchs, R. (1996), “Self-efficacy and health behaviours”, in Conner, M. and Norman, P. (Eds), Predicting Health Behaviour: Research and Practice with Social Cognition Models, Open University Press, Buckingham, pp. 163-196.

Sirgy, J.M. (1982), “Self-concepts in consumer behaviour: a critical review”, Journal of Consumer Research, Vol. 9 No. 3, pp. 287-300.

Smith, J.A., Michie, S., Stephenson, M. and Quarrell, O. (2002), “Risk perception and decision-making processes in candidates for genetic testing for Huntington’s disease: an interpretative phenomenological analysis”, Journal of Health Psychology, Vol. 7 No. 2, pp. 131-144.

Spiggle, S. (1994), “Analysis and interpretation of qualitative data in consumer research”, Journal of Consumer Research, Vol. 21 No. 3, pp. 491-503.

Stets, J.E. (2006), “Identity theory and emotions”, inStets, J.E. and Turner, J.H. (Eds), Handbook of the Sociology of Emotions, Springer US, New York, NY, pp. 203-223.

Stokes, D. (1996), “Translating social ecological theory into guidelines for community health promotion”, American Journal of Health Promotion, Vol. 10 No. 4, pp. 282-298.

Strauss, A. and Corbin, J. (1998), Basics of Qualitative Research Techniques and Procedures for Developing Grounded Theory, 2nd ed., Sage Publications, London.

Strong, K.G. and Huon, G.F. (1999), “Controlled motivation and the persistence of weight-loss dieting”, European Eating Disorders Review, Vol. 7 No. 2, pp. 136-146.

Stryker, S. and Burke, P.J. (2000), “The past, present, and future of an identity theory”, Social Psychology Quarterly, Vol. 63 No. 4, pp. 284-297.

Sundquist, J. and Johansson, S.E. (1998), “The influence of socioeconomic status, ethnicity and lifestyle on body mass index in a longitudinal study”, International Journal of Epidemiology, Vol. 27 No. 1, pp. 57-63.

Tabachnick, B.G., Fidell, L.S. and Osterlind, S.J. (2001), Using Multivariate Statistics, 4th ed., Allyn and Bacon, Boston, MA.

Tajfel, H. and Turner, J.C. (1979), “An integrative theory of intergroup conflict”, inAustin, W.G. and Worchel, S. (Eds), The Social Psychology of Intergroup Relations, Brooks/Cole, Monterey, CA, pp. 33-47.

Tajfel, H. and Turner, J.C. (1986), “The social identity theory of intergroup behaviour”, inWorchel, S. and Austin, W.G. (Eds), Psychology of Intergroup Relations, Nelson-Hall, Chicago, IL, pp. 7-24.

Taylor, S.J. and Bogdan, R. (1998), Introduction to Qualitative Research Methods: The Search for Meaning, 3rd ed., Wiley, New York, NY.

Thompson, W. and Hickey, J. (2005), Society in Focus, Pearson, Boston, MA.

Tourangeau, R. (1984), “Cognitive science and survey methods”, inThomas, B.J., Miron, L.S., Judith, M.T. and Roger, T. (Eds), Cognitive Aspects of Survey Methodology: Building a Bridge between Disciplines, National Academic Press, Washington, DC, pp. 73-100.

Trogdon, J.G., Nonnemaker, J. and Pais, J. (2008), “Peer effects in adolescent overweight”, Journal of Health Economics, Vol. 27 No. 5, pp. 1388-1399.

van den Hoek, D.J., Miller, C.T., Fraser, S.F., Selig, S.E. and Dixon, J.B. (2017), “Does exercise training augment improvements in quality of life induced by energy restriction for obese populations? A systematic review”, Quality of Life Research, Vol. 26 No. 10, pp. 2593-2605.

Vandevijvere, S. and Swinburn, B. (2015), “Getting serious about protecting New Zealand children against unhealthy food marketing”, The New Zealand Medical Journal, Vol. 128 No. 1417, pp. 36-40.

Wansink, B. (2004), “Environmental factors that increase the food intake and consumption”, Annual Review of Nutrition, Vol. 24 No. 1, pp. 455-479.

Wigfield, A. and Eccles, J.S. (2000), “Expectancy–value theory of achievement motivation”, Contemporary Educational Psychology, Vol. 25 No. 1, pp. 68-81.

Further reading

Walpole, S.C., Prieto-Merino, D., Edwards, P., Cleland, J., Stevens, G. and Roberts, I. (2012), “The weight of nations: an estimation of adult human biomass”, BMC Public Health, Vol. 12 No. 1, p. 1.

World Health Organization (2018), “Controlling the global obesity epidemic”, available at: www.who.int/nutrition/topics/obesity/en/ (accessed 15 January 2016).

Corresponding author

Sandra D. Smith can be contacted at: sd.smith@auckland.ac.nz