Modeling social norms and social influence in obesity

Converging perspectives on the processes exacerbating adolescent obesity: An integrative systems approach

Adolescent obesity is a complex public health challenge with steadily increasing and variable prevalence among countries. This paper synthesises the driving feedback mechanisms of adolescent obesity studied in the CO-CREATE project, furthering our understanding of the complexity of this issue. Using systems thinking principles and causal loop diagramming, we integrated the following knowledge and perspectives derived from diverse sources into a causal loop diagram (CLD): a systems map generated by adolescents through participatory modelling workshops, a comprehensive literature review, and input from subject-matter experts during validation workshops. We used a structured and iterative approach to include drivers and to identify feedback loops exacerbating adolescent obesity. The CLD identified 27 key feedback loops across four themes: twelve related to the commercial food environment, six to the physical activity environment, four to mental wellbeing and five to social norms. These loops indicate not only diet and physical activity as drivers of obesity but also stress and other emotional and social pressures. Recognising the imperative need to integrate the perspectives and experiences of adolescents into our analysis, this work advocates for the synthesis of experiential insights with empirical research. The integrated CLD can be used as a visual tool that fosters collaboration among stakeholders and engenders a more comprehensive and inclusive system understanding that can provide holistic intervention considerations to tackle adolescent obesity. Additionally, the CLD lays a foundation for subsequent quantitative modelling works to further address this issue and develop context-based approaches to prevention and evaluation of adolescent obesity.

Heterogeneous risk tolerance, in-groups, and epidemic waves

There is a growing interest in the joint modeling of the dynamics of disease and health-related beliefs and attitudes, but coupling mechanisms are yet to be understood. We introduce a model where risk information, which can be delayed, comes in two flavors, including historical risk derived from perceived incidence data and predicted risk information. Our model also includes an interpretation domain where the behavioral response to risk information is subject to in-group pressure. We then simulate how the strength of behavioral reaction impacts epidemic severity as measured by epidemic peak size, number of waves, and final size. Simulated behavioral response is not effective when the level of protection that prophylactic behavior provides is as small as 50% or lower. At a higher level of 75% or more, we see the emergence of multiple epidemic waves. In addition, simulations show that different behavioral response profiles can lead to various epidemic outcomes that are non-monotonic with the strength of reaction to risk information. We also modeled heterogeneity in the response profile of a population and find they can lead to less severe epidemic outcome in terms of peak size.

Academic performance and social networks of adolescents in a caribbean city in Colombia

BACKGROUND

Social factors and networks of friends can influence an adolescent's behavior, including academic performance (AP) in school. This study aimed to analyze the relationship between AP and adolescents' social networks in a Caribbean city in Colombia.

METHODS

A cross-sectional study was carried out with 806 schoolchildren from 12 to 17 years old of both sexes (52.7% girls), selected by multi-stage sampling from schools in the rural and urban areas of the city of Montería, Colombia. The AP was obtained from the school records; the sociodemographic variables included the location of the school (rural or urban), family structure, family functioning (Apgar score), and family affluence scale. Social network variables included social activity, popularity, reciprocity, homophily, friends' academic performance, network size, network density, cluster of friends, and centrality.

RESULTS

The AP was inversely associated with the Apgar score in boys. No associations of AP with the school location, family structure, family affluence scale, and age were observed. In social network variables, AP was positively associated with popularity and friends' academic performance in girls and boys, and negatively associated with homophily in boys.

CONCLUSIONS

AP was associated with social network variables. These results could help implement interventions to improve adolescents' social environment and AP.

Adolescent relational behaviour and the obesity pandemic: A descriptive study applying social network analysis and machine learning techniques

AIM

To study the existence of subgroups by exploring the similarities between the attributes of the nodes of the groups, in relation to diet and gender and, to analyse the connectivity between groups based on aspects of similarities between them through SNA and artificial intelligence techniques.

METHODS

235 students from 5 different educational centres participate in this study between March and December 2015. Data analysis carried out is divided into two blocks: social network analysis and unsupervised machine learning techniques. As for the social network analysis, the Girvan-Newman technique was applied to find the best number of cohesive groups within each of the friendship networks of the different classes analysed.

RESULTS

After applying Girvan-Newman in the three classes, the best division into clusters was respectively 2 for classroom A, 7 for classroom B and 6 for classroom C. There are significant differences between the groups and the gender and diet variables. After applying K-means using population diet as an input variable, a K-means clustering of 2 clusters for class A, 3 clusters for class B and 3 clusters for class C is obtained.

CONCLUSION

Adolescents form subgroups within their classrooms. Subgroup cohesion is defined by the fact that nodes share similarities in aspects that influence obesity, they share attributes related to food quality and gender. The concept of homophily, related to SNA, justifies our results. Artificial intelligence techniques together with the application of the Girvan-Newman provide robustness to the structural analysis of similarities and cohesion between subgroups.

Measurement accuracy and reliability of self-reported versus measured weight and height among adults in Malaysia: Findings from a nationwide blood pressure screening programme

Most studies reporting prevalence of obesity use actual weight and height measurements. Self-reported weight and height have been used in epidemiological studies as they have been shown to be reliable, convenient, and inexpensive alternatives to actual measurements. However, the accuracy of self-reported weight and height might vary in different regions because of the difference in health awareness and social influences. This study aims to determine the accuracy and reliability of self-reported weight and height compared to actual measured weight and height among adults in Malaysia. This was a cross-sectional study conducted at the community level during blood pressure screening campaigns. Participants self-reported their weight and height in a questionnaire survey. Their weight and height were validated using measurements by researchers on the same setting. Body mass index (BMI) was defined as underweight (<18.5kg/m2), normal (18.5-22.9 kg/m2), overweight (23-27.4 kg/m2) and obesity (≥27.5 kg/m2). Bland-Altman analysis, intraclass correlation coefficients and weighted Kappa statistics were used to assess the degree of agreement between self-reported and measured weight and height. A total of 2781 participants were recruited in this study. The difference between the mean self-reported and measured weight and height were 0.4 kg and 0.4 cm respectively. Weighted Kappa statistics analysis showed that there was a substantial agreement between the BMI classifications derived from self-reported and actual measurement (ҡ = 0.920, p<0.001). There was no marked difference in the sensitivity and specificity of self-reported BMI among Malaysian adults by gender. We observed substantial agreement between self-reported and measured body weight and height within a sample of Malaysian adults. While self-reported body weight showed weaker agreement with actual measurements particularly for obese and overweight individuals, BMI values derived from self-reported weight and height were accurate for 88.53% of the participants. We thus conclude that self-reported height and weight measures may be useful for tracking and estimating population trends amongst Malaysian adults.

Trends in the distribution of body mass index, waist circumference and prevalence of obesity among Taiwanese adults, 1993–2016

Background

Differences in the prevalence of general and abdominal obesity by subgroups such as age, sex, and education have been reported worldwide. Most studies in Taiwan regarding obesity prevalence were targeted at school-aged children or without further stratification by subgroups. Our aim was to examine the age-specific secular trend of body mass index (BMI), waist circumference (WC), and obesity prevalence stratified by sex, education and urbanization levels in Taiwanese adults.

Methods

We used three waves of nationally representative population from the Nutrition and Health Survey in Taiwan (NAHSIT) 1993–1996 (n = 2 989), 2005–2008 (n = 2 495), and 2013–2016 (n = 2 880). The data included standardized measurement of body weight, height, and WC. We conducted a serial cross-sectional analysis among adults aged 20 years or above to examine the age-specific trends of BMI, WC, and the prevalence of underweight, overweight, general obesity, and abdominal obesity with stratification by sex, education, and urbanization levels.

Results

The general obesity prevalence was 16%, 21%, and 20% and the abdominal obesity prevalence was 27%, 42%, and 47% in the 1993–1996, 2005–2008, and 2013–2016 surveys, respectively. The age-specific secular trend of BMI differed across subgroups; however, the trend of WC increased rapidly regardless of subgroups, except for women aged ≥60 years. The general obesity prevalence increased noticeably among men, younger- and middle-age adults with high school or higher education, middle- and older-age adults with lower than high school education, people <39 and ≥50 years of age residing in rural areas, and among those between 30 and 59 and ≥70 years of age residing in urban areas.

Conclusions

Although the increasing trend of general obesity prevalence was levelling off among several subgroups, the abdominal obesity prevalence increased significantly and rapidly in Taiwan. Future research in developing effective weight and WC control interventions tailored to different subgroups is urgently needed.

Relation of social network support to child health behaviors among children in treatment for overweight/obesity

PURPOSE

Little is known about the influence of social network support on child health behaviors in the context of weight-loss interventions. This study examined the associations between a child's co-participation (i.e., network support) in weight-related health behaviors (i.e., physical and sedentary activity, eating behavior) and the child's own health behaviors during family-based behavioral treatment (FBT).

METHODS

Children (n = 241) with overweight/obesity (mean age = 9.4 ± 1.3y; 63% female) completed semi-structured interviews assessing network support for healthy/unhealthy eating and physical/sedentary activity, and a 3-day dietary recall. Physical activity was assessed with accelerometry, and sedentary activity was measured via parent-reported child screen time use. All assessments were taken at baseline and after 4 months of FBT. Hierarchical linear regressions examined changes in network support as they related to changes in health behaviors from baseline to the end of FBT.

RESULTS

Changes in network support for healthy eating were related to changes in vegetable, but not fruit, intake across FBT, while changes in network support for unhealthy eating were negatively related to changes in diet quality. Changes in network support for sedentary activity were negatively related to changes in minutes of physical activity and positively related to changes in screen time.

CONCLUSION

The present findings suggest that a child's network support for health behaviors may relate to behavior change among children during FBT and provide opportunities for targeted intervention.

LEVEL OF EVIDENCE

III. cohort study.

Does Self-Perception Equal the Truth When Judging Own Body Weight and Height?

Background: Data from the research project “Fitness clubs—a venue for public health?” provided an opportunity to evaluate the accuracy of self-reported body weight and height, and subsequent Body Mass Index (BMI), as well as the “trueness” of novice exercisers perception of weight status category, which has not been examined in this population. The aims were to examine self-reported body weight, height, and calculated BMI data from an online survey compared with measured data at fitness club start-up, investigate how accurately novice exercisers place themselves within self-classified weight group (underweight, normal weight, overweight, and obese), and compare this with fitness club attendance at three months follow-up. Methods: Prior to anthropometric measurements, 62 men and 63 women responded to an online questionnaire, including body weight (kilogram, kg) and height (centimeters, cm), and self-classified weight group (“I think I am … underweight, normal weight, overweight, obese”). We used the following statistical analysis: Paired sample t-tests, a Bland–Altman plot kappa statistics, chi-squared tests, and a logistic regression. Results: Mean difference of BMI calculated from self-reported and measured data was 0.06 (95% CI −0.29 to 0.17, p = 0.593) in men, and 0.16 (95% CI −0.40 to 0.09, p = 0.224) in women, with four participants being outliers of the 95% limits of agreement (Bland-Altman plot). Allowing a difference of 0.5 kg between self-reported and measured weight, we found that 16% reported their weight correctly, 31.2% underreported (−1.89 ± 1.59 kg), and 52.8% overreported (1.85 ± 1.23 kg), with no sex differences (p = 0.870). Further, our results suggest that both sexes may have difficulty recognizing overweight/obesity in themselves, and particularly men are likely to underreport their perceived weight group compared with women. More than half (53.3%) of the overweight men perceived themselves to be normal weight (women: 14%), and only 33.3% of obese men and women correctly classified themselves as being obese. We did not find any difference between participants correctly or incorrectly classifying weight group and fitness club attendance (≥2 times a week) at three months follow-up. Conclusion: Both sexes reported body weight and height reasonably accurately, and BMI based on self-report appears to be valid measure. Still, a large proportion of novice exercisers do not recognise their own overweight or obesity status, which may in part explain why public health campaigns do not reach risk populations.

Social Networks and Chronic Illness Management among Low-Income Tenants in Publicly Subsidized Housing: Findings from a Pilot Study

Low-income tenants in publicly subsidized housing (PSH) have higher rates of chronic illnesses than non-PSH-based residents, making the implementation of chronic illness management (CIM) essential. Based on the person-in-environment framework used in social work practice, which emphasizes the importance of interactions between clients and their social environment, this pilot study used personal network analysis (PNA), a variant of social network analysis, to explore what attributes of social networks are relevant to CIM among 26 low-income tenants independently living in a PSH in the Southwest United States. Tenants with a smaller network size and effective size presented better self-efficacy to manage chronic disease (SEMCD) and lower levels of depressive symptoms. Being connected to a higher proportion of alters (i.e., network members) with whom they discussed health matters was also associated with high SEMCD. As for CIM implementation, being connected to a doctor and a lower proportion of alters who have chronic illnesses were associated with doing aerobic exercise. Tenants with a larger proportion of alters serving multiple functions reported more frequent vegetable and fruit consumption, while those with a higher share of kin alters and lower share of alters living in the same PSH reported less frequent high-fat food consumption. Our findings help social workers discover relevant social networks and dynamics that low-income tenants at PSH capitalize to locate resources for CIM. Further studies are recommended to adopt PNA to expand practice-related knowledge that social workers can use for health promotion among low-income tenants with chronic illness.

Introduction to the Theme Issue on Dynamics of Health Behavior: Revisiting Systems Science for Population Health

Systems science can help public health professionals to better understand the complex dynamics between factors affecting health behaviors and outcomes and to identify intervention opportunities. Despite their demonstrated utility in addressing health topics such influenza, tobacco control, and obesity, the associated methods continue to be underutilized by researchers and practitioners addressing health behaviors. This article discusses the growth of systems science methods (e.g., system dynamics, social network analysis, and agent-based modeling) in health research, provides a frame for the articles included in this themed issue, and closes with recommendations for enhancing the future of systems science and health behavior research. We argue that integrating systems sciences methods into health behavior research and practice is essential for improved population health and look forward to supporting the evolution of the field.

Design and methods of Shape Up Under 5: Integration of systems science and community-engaged research to prevent early childhood obesity

Shape Up Under 5 (SUU5) was a two-year early childhood obesity prevention pilot study in Somerville, Massachusetts (2015-2017) designed to test a novel conceptual framework called Stakeholder-driven Community Diffusion. For whole-of-community interventions, this framework posits that diffusion of stakeholders' knowledge about and engagement with childhood obesity prevention efforts through their social networks will improve the implementation of health-promoting policy and practice changes intended to reduce obesity risk. SUU5 used systems science methods (agent-based modeling, group model building, social network analysis) to design, facilitate, and evaluate the work of 16 multisector stakeholders ('the Committee'). In this paper, we describe the design and methods of SUU5 using the conceptual framework: the approach to data collection, and methods and rationale for study inputs, activities and evaluation, which together may further our understanding of the hypothesized processes within Stakeholder-driven Community Diffusion. We also present a generalizable conceptual framework for addressing childhood obesity and similar complex public health issues through whole-of-community interventions.

Assessing the Impact of Lifestyle Interventions on Diabetes Prevention in China: A Modeling Approach

China's diabetes epidemic is getting worse. People with diabetes in China usually have a lower body weight and a different lifestyle profile compared to their counterparts in the United States (US). More and more evidence show that certain lifestyles can possibly be spread from person to person, leading some to propose considering social influence when establishing preventive policies. This study developed an innovative agent-based model of the diabetes epidemic for the Chinese population. Based on the risk factors and related complications of diabetes, the model captured individual health progression, quantitatively described the peer influence of certain lifestyles, and projected population health outcomes over a specific time period. We simulated several hypothetical interventions (i.e., improving diet, controlling smoking, improving physical activity) and assessed their impact on diabetes rates. We validated the model by comparing simulation results with external datasets. Our results showed that improving physical activity could result in the most significant decrease in diabetes prevalence compared to improving diet and controlling smoking. Our model can be used to inform policymakers on how the diabetes epidemic develops and help them compare different diabetes prevention programs in practice.

Proximity effects in obesity rates in the US: A Spatial Markov Chains approach

In this paper, we investigate, by means of a Spatial Markov Chains approach, the existence of proximity effects at State level for US data on obesity rates in the period 1990-2011. We find that proximity effects do play an important role in the spatial diffusion of obesity (the obesity 'epidemics'), and that the actual health geography of nearby States in terms of high vs. low obesity rates makes an important difference as to the future evolution of the State's own obesity rate over time. This means, in particular, that clusters of States characterized by uniformly high levels of obesity rates, as it happens for instance in the US Southern macro-region, may suffer from a perverse 'geographical lock-in' effect that calls for coordinated action across States to implement effective countervailing policies.

From awareness to behaviour: Testing a hierarchy of effects model on the Australian Make Healthy Normal campaign using mediation analysis

The Make Healthy Normal mass media campaign was a three-year campaign launched in 2015 in New South Wales (NSW), Australia to address community norms around overweight and obesity. It was underpinned by a hierarchy of effects model; a commonly used framework in campaigns but one that has rarely been tested. The campaign evaluation included a cohort study of NSW adults, surveyed three times over 12 months (n = 939 at Wave 3). This study tested the campaign's hierarchy of effects model, which theorized that participants would move from recognition to behaviour change via understanding, knowledge, attitude, social norms, self-efficacy, and intention, using these data. We used the moderation and mediation of effects method proposed by Baron and Kenny, adjusting for age and sex, to test for progression through the hierarchy of effects for two outcomes: physical activity and fast food consumption. We found a clear progression through the theorized model, from recognition through to behaviour change, via the intermediate variables for both outcomes. We also found several effects not predicted by the theorized model, with consistently strong associations between understanding and attitude, understanding and self-efficacy, attitude and self-efficacy, and self-efficacy and behaviour change in both outcome models. Our study provides support for the hierarchy of effects as a conceptual model in campaign planning and evaluation of social marketing campaigns. To our knowledge, this is the first study to compare the hierarchy between two behavioural outcomes and the consistency observed between the models adds to the potential usefulness of the hierarchy of effects.

A comparison of measured versus self-reported anthropometrics for assessing obesity in adults: a literature review

Aims: Up-to-date information on the accuracy between different anthropometric data collection methods is vital for the reliability of anthropometric data. A previous review on this matter was conducted a decade ago. Our aim was to conduct a literature review on the accuracy of self-reported height, weight, and body mass index (BMI) against measured values for assessing obesity in adults. To obtain an overview of the present situation, we included studies published after the previous review. Differences according to sex, BMI groups, and continents were also assessed. Methods: Studies published between January 2006 and April 2017 were identified from a literature search on PubMed. Results: Our search retrieved 62 publications on adult populations that showed a tendency for self-reported height to be overestimated and weight to be underestimated when compared with measured values. The findings were similar for both sexes. BMI derived from self-reported height and weight was underestimated; there was a clear tendency for underestimation of overweight (from 1.8%-points to 9.8%-points) and obesity (from 0.7%-points to 13.4%-points) prevalence by self-report. The bias was greater in overweight and obese participants than those of normal weight. Studies conducted in North America showed a greater bias, whereas the bias in Asian studies seemed to be lower than those from other continents. Conclusions: With globally rising obesity rates, accurate estimation of obesity is essential for effective public health policies to support obesity prevention. As self-report bias tends to be higher among overweight and obese individuals, measured anthropometrics provide a more reliable tool for assessing the prevalence of obesity.

Dynamics of Implementation and Maintenance of Organizational Health Interventions

In this study, we present case studies to explore the dynamics of implementation and maintenance of health interventions. We analyze how specific interventions are built and eroded, how the building and erosion mechanisms are interconnected, and why we can see significantly different erosion rates across otherwise similar organizations. We use multiple comparative obesity prevention case studies to provide empirical information on the mechanisms of interest, and use qualitative systems modeling to integrate our evolving understanding into an internally consistent and transparent theory of the phenomenon. Our preliminary results identify reinforcing feedback mechanisms, including design of organizational processes, motivation of stakeholders, and communication among stakeholders, which influence implementation and maintenance of intervention components. Over time, these feedback mechanisms may drive a wedge between otherwise similar organizations, leading to distinct configurations of implementation and maintenance processes.

The dual role of friendship and antipathy relations in the marginalization of overweight children in their peer networks: The TRAILS Study

Weight-based stigma compromises the social networks of overweight children. To date, research on the position of overweight children in their peer network has focused only on friendship relations, and not on negative relationship dimensions. This study examined how overweight was associated with relations of friendship and dislike (antipathies) in the peer group. Exponential random graph models (ERGM) were used to examine friendship and antipathy relations among overweight children and their classmates, using a sub-sample from the TRacking Adolescents’ Individual Lives Survey (N = 504, M age 11.4). Findings showed that overweight children were less likely to receive friendship nominations, and were more likely to receive dislike nominations. Overweight children were also more likely than their non-overweight peers to nominate classmates that they disliked. Together, the results indicate that positive and negative peer relations are impacted by children’s weight status, and are relevant to addressing the social marginalization of overweight children.

A multifactorial obesity model developed from nationwide public health exposome data and modern computational analyses

STATEMENT OF THE PROBLEM

Obesity is both multifactorial and multimodal, making it difficult to identify, unravel and distinguish causative and contributing factors. The lack of a clear model of aetiology hampers the design and evaluation of interventions to prevent and reduce obesity.

METHODS

Using modern graph-theoretical algorithms, we are able to coalesce and analyse thousands of inter-dependent variables and interpret their putative relationships to obesity. Our modelling is different from traditional approaches; we make no a priori assumptions about the population, and model instead based on the actual characteristics of a population. Paracliques, noise-resistant collections of highly-correlated variables, are differentially distilled from data taken over counties associated with low versus high obesity rates. Factor analysis is then applied and a model is developed.

RESULTS AND CONCLUSIONS

Latent variables concentrated around social deprivation, community infrastructure and climate, and especially heat stress were connected to obesity. Infrastructure, environment and community organisation differed in counties with low versus high obesity rates. Clear connections of community infrastructure with obesity in our results lead us to conclude that community level interventions are critical. This effort suggests that it might be useful to study and plan interventions around community organisation and structure, rather than just the individual, to combat the nation's obesity epidemic.

Obesity as self-regulation failure: A “disease of affluence” that selectively hits the less affluent?

An effect of the long-term cycle of industrial and post-industrial global development is the increasingly generalized access to abundant and diversified food sources. This poses a substantial problem of self-regulation that mainly affects the less affluent and whose failures may play an important role in the explanation of the increasing social incidence of overweight and obesity problems.

Using Agent-Based Models to Develop Public Policy about Food Behaviours: Future Directions and Recommendations

Most adults are overweight or obese in many western countries. Several population-level interventions on the physical, economical, political, or sociocultural environment have thus attempted to achieve a healthier weight. These interventions have involved different weight-related behaviours, such as food behaviours. Agent-based models (ABMs) have the potential to help policymakers evaluate food behaviour interventions from a systems perspective. However, fully realizing this potential involves a complex procedure starting with obtaining and analyzing data to populate the model and eventually identifying more efficient cross-sectoral policies. Current procedures for ABMs of food behaviours are mostly rooted in one technique, often ignore the food environment beyond home and work, and underutilize rich datasets. In this paper, we address some of these limitations to better support policymakers through two contributions. First, via a scoping review, we highlight readily available datasets and techniques to deal with these limitations independently. Second, we propose a three steps' process to tackle all limitations together and discuss its use to develop future models for food behaviours. We acknowledge that this integrated process is a leap forward in ABMs. However, this long-term objective is well-worth addressing as it can generate robust findings to effectively inform the design of food behaviour interventions.

Protocol for a randomized controlled trial testing the impact of feedback on familial risk of chronic diseases on family-level intentions to participate in preventive lifestyle behaviors

Background

Common disease risk clusters in families due to shared genetics, exposure to environmental risk factors, and because many health behaviours are established and maintained in family environments. This randomised controlled trial will test whether the provision of a family health history (FHH) risk assessment tool increases intentions and engagement in health behaviors. Message distribution and collective behavior change within family networks will be mapped using social network analysis. The relative intervention impact will be compared between families from different ethnic backgrounds.

Methods

One hundred and fifty mothers (50 Anglo-Australian, 50 Italian-Australian, 50 Vietnamese-Australian) will be recruited, with four or more other family members across three generations, including a child (aged 10–18 years). Each family is randomly assigned to intervention or control. At baseline and 6-month follow-up, all participants complete surveys to assess dietary and physical activity intentions and behaviors, attitudes towards food, and perceived disease risk. Intervention families receive a visual pedigree detailing their FHH of diabetes, heart disease, breast and bowel cancer, a health education workbook to ascertain members’ disease risk (i.e. average or above average risk), and screening and primary prevention recommendations. After completion of follow-up assessments, controls will receive their pedigree and workbook. The primary hypothesis is that attitudes and lifestyle behaviors will improve more within families exposed to FHH feedback, although the extent of this improvement may vary between families from different ethnic backgrounds. Additionally, the extent of improvement in the treatment group will be moderated by the level of family disease risk, with above-average risk leading to greater improvement. A secondary aim will explore different family members’ roles in message distribution and collective responses to risk using social network approaches and to compare network functioning between families with different ethnic backgrounds.

Discussion

Results will guide future health promotion programs aimed at improving lifestyle factors. This research will assess whether FHH can motivate families to adopt family-level strategies to support health promoting behaviors. Secondary analyses aim to identify change agents within the family who are particularly effective in shifting normative behaviors.

Trial registration

Australian New Zealand Clinical Trials Registry ACTRN12613001033730. Retrospectively registered: 17 September, 2013.

Nutrition and the science of disease prevention: a systems approach to support metabolic health

Progress in nutritional science, genetics, computer science, and behavioral economics can be leveraged to address the challenge of noncommunicable disease. This report highlights the connection between nutrition and the complex science of preventing disease and discusses the promotion of optimal metabolic health, building on input from several complementary disciplines. The discussion focuses on (1) the basic science of optimal metabolic health, including data from gene-diet interactions, microbiome, and epidemiological research in nutrition, with the goal of defining better targets and interventions, and (2) how nutrition, from pharma to lifestyle, can build on systems science to address complex issues.