Big Data and Systems Methods: The Next Frontier to Tackling the Global Obesity Epidemic

Non-traditional data sources in obesity research: a systematic review of their use in the study of obesogenic environments

BACKGROUND

The complex nature of obesity increasingly requires a comprehensive approach that includes the role of environmental factors. For understanding contextual determinants, the resources provided by technological advances could become a key factor in obesogenic environment research. This study aims to identify different sources of non-traditional data and their applications, considering the domains of obesogenic environments: physical, sociocultural, political and economic.

METHODS

We conducted a systematic search in PubMed, Scopus and LILACS databases by two independent groups of reviewers, from September to December 2021. We included those studies oriented to adult obesity research using non-traditional data sources, published in the last 5 years in English, Spanish or Portuguese. The overall reporting followed the PRISMA guidelines.

RESULTS

The initial search yielded 1583 articles, 94 articles were kept for full-text screening, and 53 studies met the eligibility criteria and were included. We extracted information about countries of origin, study design, observation units, obesity-related outcomes, environment variables, and non-traditional data sources used. Our results revealed that most of the studies originated from high-income countries (86.54%) and used geospatial data within a GIS (76.67%), social networks (16.67%), and digital devices (11.66%) as data sources. Geospatial data were the most utilised data source and mainly contributed to the study of the physical domains of obesogenic environments, followed by social networks providing data to the analysis of the sociocultural domain. A gap in the literature exploring the political domain of environments was also evident.

CONCLUSION

The disparities between countries are noticeable. Geospatial and social network data sources contributed to studying the physical and sociocultural environments, which could be a valuable complement to those traditionally used in obesity research. We propose the use of information available on the Internet, addressed by artificial intelligence-based tools, to increase the knowledge on political and economic dimensions of the obesogenic environment.

The Need for Systems Approaches for Precision Communications in Public Health

A major challenge in communicating health-related information is the involvement of multiple complex systems from the creation of the information to the sources and channels of dispersion to the information users themselves. To date, public health communications approaches have often not adequately accounted for the complexities of these systems to the degree necessary to have maximum impact. The virality of COVID-19 misinformation and disinformation has brought to light the need to consider these system complexities more extensively. Unaided, it is difficult for humans to see and fully understand complex systems. Luckily, there are a range of systems approaches and methods, such as systems mapping and systems modeling, that can help better elucidate complex systems. Using these methods to better characterize the various systems involved in communicating public health-related information can lead to the development of more tailored, precise, and proactive communications. Proceeding in an iterative manner to help design, implement, and adjust such communications strategies can increase impact and leave less opportunity for misinformation and disinformation to spread.

The Impact of a Place-Tailored Digital Health App Promoting Exercise Classes on African American Women's Physical Activity and Obesity: Simulation Study

BACKGROUND

The increasing prevalence of smartphone apps to help people find different services raises the question of whether apps to help people find physical activity (PA) locations would help better prevent and control having overweight or obesity.

OBJECTIVE

The aim of this paper is to determine and quantify the potential impact of a digital health intervention for African American women prior to allocating financial resources toward implementation.

METHODS

We developed our Virtual Population Obesity Prevention, agent-based model of Washington, DC, to simulate the impact of a place-tailored digital health app that provides information about free recreation center classes on PA, BMI, and overweight and obesity prevalence among African American women.

RESULTS

When the app is introduced at the beginning of the simulation, with app engagement at 25% (eg, 25% [41,839/167,356] of women aware of the app; 25% [10,460/41,839] of those aware downloading the app; and 25% [2615/10,460] of those who download it receiving regular push notifications), and a 25% (25/100) baseline probability to exercise (eg, without the app), there are no statistically significant increases in PA levels or decreases in BMI or obesity prevalence over 5 years across the population. When 50% (83,678/167,356) of women are aware of the app; 58.23% (48,725/83,678) of those who are aware download it; and 55% (26,799/48,725) of those who download it receive regular push notifications, in line with existing studies on app usage, introducing the app on average increases PA and decreases weight or obesity prevalence, though the changes are not statistically significant. When app engagement increased to 75% (125,517/167,356) of women who were aware, 75% (94,138/125,517) of those who were aware downloading it, and 75% (70,603/94,138) of those who downloaded it opting into the app's push notifications, there were statistically significant changes in PA participation, minutes of PA and obesity prevalence.

CONCLUSIONS

Our study shows that a digital health app that helps identify recreation center classes does not result in substantive population-wide health effects at lower levels of app engagement. For the app to result in statistically significant increases in PA and reductions in obesity prevalence over 5 years, there needs to be at least 75% (125,517/167,356) of women aware of the app, 75% (94,138/125,517) of those aware of the app download it, and 75% (70,603/94,138) of those who download it opt into push notifications. Nevertheless, the app cannot fully overcome lack of access to recreation centers; therefore, public health administrators as well as parks and recreation agencies might consider incorporating this type of technology into multilevel interventions that also target the built environment and other social determinants of health.

Machine learning to predict in‐hospital mortality among patients with severe obesity: Proof of concept study

Objectives

Hospitalized patients with severe obesity require adapted hospital management. The aim of this study was to evaluate a machine learning model to predict in‐hospital mortality among this population.

Methods

Data of unselected consecutive emergency department admissions of hospitalized patients with severe obesity (BMI ≥ 40 kg/m²) was analyzed. Data was retrieved from five hospitals from the Mount Sinai health system, New York. The study time frame was between January 2011 and December 2019. Data was used to train a gradient‐boosting machine learning model to identify in‐hospital mortality. The model was trained and evaluated based on the data from four hospitals and externally validated on held‐out data from the fifth hospital.

Results

A total of 14,078 hospital admissions of inpatients with severe obesity were included. The in‐hospital mortality rate was 297/14,078 (2.1%). In univariate analysis, albumin (area under the curve [AUC] = 0.77), blood urea nitrogen (AUC = 0.76), acuity level (AUC = 0.73), lactate (AUC = 0.72), and chief complaint (AUC = 0.72) were the best single predictors. For Youden’s index, the model had a sensitivity of 0.77 (95% CI: 0.67–0.86) with a false positive rate of 1:9.

Conclusion

A machine learning model trained on clinical measures provides proof of concept performance in predicting mortality in patients with severe obesity. This implies that such models may help to adopt specific decision support tools for this population.