Network approach in health behavior research: how can we explore new questions?

Psychological network analysis of the relations between beliefs about smoking for smokers with and without a recent quit attempt

OBJECTIVE

Psychological network analysis was used to evaluate the relations between beliefs about cigarette smoking in current smokers with and without a recent quit attempt and determine if these networks differed in global strength (how strongly beliefs are related) or global structure (which beliefs are related).

DESIGN

Using two publicly available datasets, the California Smokers' Cohort (CSC; N = 933) and Wave 1 of the Population Assessment of Tobacco and Health (PATH; N = 7855), we evaluated differences in global strength and global structure of the beliefs held by current smokers with and without a recent quit attempt.

MAIN OUTCOME MEASURES

Strength and structure of networks generated for current smokers with and without a recent quit attempt.

RESULTS

In the CSC dataset there were differences between smokers with and without a recent quit attempt in global structure and marginal differences in global strength; the PATH dataset suggested small differences in global strength and structure.

CONCLUSION

The present study suggests that consideration of relations among smoking beliefs may be a valuable contribution to characterizing smoking beliefs when assessing smoking quit attempts.

A Note on Ising Network Analysis with Missing Data

The Ising model has become a popular psychometric model for analyzing item response data. The statistical inference of the Ising model is typically carried out via a pseudo-likelihood, as the standard likelihood approach suffers from a high computational cost when there are many variables (i.e., items). Unfortunately, the presence of missing values can hinder the use of pseudo-likelihood, and a listwise deletion approach for missing data treatment may introduce a substantial bias into the estimation and sometimes yield misleading interpretations. This paper proposes a conditional Bayesian framework for Ising network analysis with missing data, which integrates a pseudo-likelihood approach with iterative data imputation. An asymptotic theory is established for the method. Furthermore, a computationally efficient Pólya-Gamma data augmentation procedure is proposed to streamline the sampling of model parameters. The method's performance is shown through simulations and a real-world application to data on major depressive and generalized anxiety disorders from the National Epidemiological Survey on Alcohol and Related Conditions (NESARC).

Symptom cluster among cancer survivors from a nationally representative survey: a network analysis

PURPOSE

To identify the symptom cluster among cancer survivors and examine their subgroup differences via network analysis based on nationally representative data.

METHODS

This cross-sectional study included 2966 survivors participating in the 2020 National Health Interview Survey (NHIS). Participants self-reported the presence of 14 symptoms capturing four clusters (physical, somatic, sleep, and psychologic problems). Network analysis models were used to reveal the relationships between symptoms and those interactions. Network comparison tests were applied to compare subgroups.

RESULTS

The core symptoms of the symptom cluster were fatigue (Bet = 33, Clo = 0.0067, Str = 0.9397), pain (Bet = 11, Clo = 0.0060, Str = 0.9226), wake up well rested (Bet = 25, Clo = 0.0057, Str = 0.8491), and anxiety (Bet = 5, Clo = 0.0043, Str = 0.9697) among cancer survivors. The core symptoms, network structure, and global strength were invariant between time since diagnoses (< 2 years vs. ≥ 2 years) or between numbers of cancers (1 vs. ≥ 2), yet varied between the comorbidity group and non-comorbidity group (≥ 1 vs. 0).

CONCLUSIONS

Fatigue would be a potential target for alleviating other symptoms through a negative feedback loop of other related symptoms of cancer survivors. In particular, cancer survivors with other chronic diseases should be the focus of attention and strengthen targeted intervention.

Symptom networks in older adults with cancer: A network analysis

INTRODUCTION

Due to aging, older adults with cancer (OAC) may be confronted with a complex interplay of multiple age-related issues; coupled with receiving cancer treatment, OAC may experience multiple concurrent symptoms that require the identification of the core symptom for effective management. Constructing symptom networks will help in the identification of core symptoms and help achieve personalized and precise interventions. Currently, few studies have used symptom networks to identify core symptoms in OAC. Our objectives were to construct symptom networks of OAC, explore the core symptoms, and compare the differences in symptom networks among various subgroups.

MATERIALS AND METHODS

Secondary analysis was performed using data from 485 OAC collected in 2021 from a cross-sectional survey named the Shanghai CANcer Survivor (SCANS) Report. The MD Anderson Symptom Inventory (MDASI) was used to assess the incidence and severity of cancer-related symptoms. We used the R package to construct symptom networks and identify the centrality indices. The network comparison test was used to compare network differences among the subgroups.

RESULTS

The most common and severe symptoms reported were fatigue, disturbed sleep, and difficulty remembering. The network density was 0.718. Vomiting (rs = 1.81, rb = 2.13), fatigue (rs = 1.54, rb = 1.93), and sadness (rs = 0.81, rb = 0.69) showed the highest strength values, which suggested that these symptoms were more likely to co-occur with other symptoms. The network comparison tests showed significant differences in symptom network density between the subgroups categorized as survival "< 5 years" and survival "≥ 5 years" (p = 0.002), as well as between the those with comorbidities and those without comorbidities (p = 0.037).

DISCUSSION

Our study identified symptom networks in 485 OAC. Vomiting, fatigue, and sadness were important symptoms in the symptom networks of OAC. The symptom networks differed among populations with different survival durations and comorbidities. Our network analysis provides a reference for future targeted symptom management and interventions in OAC. In the future, conducting dynamic research on symptom networks will be crucial to explore interaction mechanisms and change trends between symptoms.

Symptoms experienced after transcatheter arterial chemoembolization in patients with primary liver cancer: A network analysis

Objective

This study aimed to establish a symptom network for patients with primary liver cancer posttranscatheter arterial chemoembolization (TACE), identifying core and bridge symptoms. The goal is to provide a foundation for precise and comprehensive nursing interventions.

Methods

A total of 1207 post-TACE patients were included using a consecutive sampling method. Data collection involved a general information questionnaire, the Anderson Symptom Assessment Scale, and a primary liver cancer-specific symptom module. The symptom network was constructed using the R language.

Results

In the overall network, distress exhibited the highest strength (rs = 1.31) and betweenness (rb = 62). Fatigue had the greatest closeness (rc = 0.0043), while nausea and vomiting (r = 0.76 ± 0.02) had the highest marginal weights. Nausea had the highest bridge strength (rbs = 5.263). In the first-time TACE-treated symptom network, sadness (rbs = 5.673) showed the highest bridge strength, whereas in the non-first-time symptom network, fever (rbs = 3.061) had the highest bridge strength.

Conclusions

Distress serves as a core symptom, and nausea acts as a bridge symptom after TACE treatment in liver cancer patients. Interventions targeting bridge symptoms should be tailored based on the number of treatments, enhancing the quality of symptom management.

Network structure of the links between extracurricular time-use and delinquent behaviors: Moving forward and beyond linear relations

Using psychological network analysis, this study explored the heterogeneity of the network structure between extracurricular time-use and delinquency using a nationally representative longitudinal survey of at-school students in China (N = 10,279, 47.3% female, average age 13.6, 91.2% Han ethnicity). The results are threefold: First, time stimulation of activities occurs on weekdays, while time displacement and stimulation occur on weekends. Second, delinquent behaviors are positively correlated, forming a problem behavior syndrome. Smoking or drinking is the central delinquent behavior. Third, negative consequences of specific time-use behaviors are more likely to occur on weekends than on weekdays, and time-use behavior may function differently on weekdays versus weekends. Among them, going to coffeenets or game-centers serves as the bridge with the highest potential of triggering delinquency.

A psychosocial network approach studying biomedical HIV prevention uptake between 2017 and 2019

Biomedical HIV-prevention strategies (BmPS) among men who have sex with men (MSM), such as pre-exposure prophylaxis (PrEP) and viral load sorting (VLS), are essential but relatively new and their uptake gradual. Using an extension of the causal attitude network approach, we investigated which beliefs are related to uptake of PrEP and VLS at each time-point. We included 632 HIV-negative MSM from the Amsterdam Cohort Studies from four data-waves between 2017 and 2019. We estimated weighted, undirected networks for each time-point, where we included pairwise interactions of PrEP and VLS uptake and related beliefs. PrEP use increased from 10 to 31% (p < 0.001), while VLS was reported by 7-10% at each time-point. Uptake of both BmPS was directly related to the perceived positive impact of the strategy on one's quality of sex life and perceived supportive social norms. Overall network structure differed between time points, specifically in regard to PrEP. At earlier time points, perceptions of efficacy and affordability played an important role for PrEP uptake, while more recently social and health-related concerns became increasingly important.The network structure differed across data-waves, suggesting specific time changes in uptake motives. These findings may be used in communication to increase prevention uptake.

Relationship of the Behavior of Older Participants with Body Composition Change: Results of the Second Wave of the Cognition of Older People, Education, Recreational Activities, Nutrition, Comorbidities, and Functional Capacity Studies (COPERNICUS)

BACKGROUND

To examine the relationship between the frequency of physical activities and food product consumption with body composition change after two years in a sample of older people.

METHODS

Body composition, mass change, frequency of physical activity, and food products consumption were measured. Depression severity, health self-assessment, cognitive function, and demographic data were included as confounders.

RESULTS

There were no significant changes in body composition except for a reduction in visceral fat level within two years (p < 0.05). Drinking beer and eating sweets a few times per week were associated with a significant increase in body fat percentage (p < 0.05). Drinking green or white tea more frequently than a few times per year was related to an increase in body fat (3.18 to 3.88%, p < 0.05). Contrarily, daily consumption of coffee was related to a decrease in body fat (p = 0.029). Subjects who ate sweets once a week or more frequently consumed coffee more often.

CONCLUSIONS

More frequent drinking of beer or of green or white tea and consumption of sweets were related to an increase in body fat percentage, while daily coffee consumption was related to a decrease in body fat percentage after two years in older, healthy subjects. Noteworthily, the frequencies of food product consumption are interrelated.

A Network Analysis of Multiple Preconception Health Behaviors in Chinese Women

BACKGROUND

A healthy preconception lifestyle, consisting of multiple health behaviors, is crucial for preventing adverse health outcomes in mothers and offspring. Although knowledge about the pattern of inter-behavior relations may provide insights for nudging multiple health behavior changes, this has not been adequately explored in the existing literature. Adopting a network perspective, the present study conceptualized multiple health behaviors as a behavior network (i.e., behaviors as nodes, inter-behavior relations as edges) and utilized network analysis to investigate the pattern of interrelations of preconception health behaviors in a large sample of Chinese women.

METHOD

We used the data of a population-based cohort study in China to estimate the behavior network. An analytic sample included 41,127 Chinese women who were surveyed about their adoptions of multiple health behaviors during the preconception period.

RESULTS

Network analysis revealed a relatively dense behavior network and visualized the network structure of multiple preconception health behaviors. Subsequent centrality analysis identified three central behaviors (i.e., avoiding second- or third-hand smoke, reducing psychosocial stress, and reducing alcohol) that had distinctively stronger connections to other behaviors.

CONCLUSIONS

Preconception health behaviors were strongly interconnected, and certain behaviors had stronger influences than others within the behavior network. Our findings highlight the strong inter-relatedness of preconception health behaviors. This study also encourages targeting the three central behaviors in preconception lifestyle promotions because this may bring more secondary improvements on other non-targeted behaviors and thereby achieve comprehensive lifestyle change.

Analysis of dynamic psychological processes to understand and promote physical activity behaviour using intensive longitudinal methods: a primer

Physical activity behaviour displays temporal variability, and is influenced by a range of dynamic psychological processes (e.g., affect) and shaped by various co-occurring events (e.g., social/environmental factors, interpersonal dynamics). Yet, most physical activity research tends not to examine the dynamic psychological processes implicated in adopting and maintaining physical activity. Intensive longitudinal methods (ILM) represent one particularly salient means of studying the complex psychological dynamics that underlie and result from physical activity behaviour. With the increased recent interest in using intensive longitudinal data to understand specific dynamic psychological processes, the field of exercise and health psychology is well-positioned to draw from state-of-the-art measurement and statistical approaches that have been developed and operationalised in other fields of enquiry. The purpose of this review is to provide an overview of some of the fundamental dynamic measurement and modelling approaches applicable to the study of physical activity behaviour change, as well as the dynamic psychological processes that contribute to such change.

Identifying subtypes of HIV/AIDS-related symptoms in China using latent profile analysis and symptom networks

The identification of subgroups of people living with HIV in China based on the severity of symptom clusters and individual symptoms is crucial to determine group-specific symptom management strategies. Participants reported 27 highly prevalent HIV/AIDS-related symptoms. Latent profile analysis based on symptom severity was used to identify person-centered subtypes of HIV/AIDS-related symptoms. Symptom networks were compared among subgroups identified by latent profile analysis. A total of 2927 eligible people living with HIV (PWH) were included in the analysis. Five profiles were identified: "Profile 1: all low symptom severity" (n₂ = 2094, 71.54%), "Profile 2: medium symptom severity with syndemic conditions" (n₃ = 109, 3.72%), "Profile 3: medium symptom severity with low functional status" (n₁ = 165, 5.64%), "Profile 4: medium symptom severity in transitional period" (n₄ = 448, 15.31%), and "Profile 5: all high symptom severity" (n₅ = 111, 3.79%). Except for Profile 1 and Profile 5, the symptom severity was similar among the other three profiles. Profiles 1 (2.09 ± 0.52) and 4 (2.44 ± 0.66) had the smallest ∑s values, and Profiles 2 (4.38 ± 1.40) and 5 (4.39 ± 1.22) had the largest ∑s values. Our study demonstrates the need for health care professionals to provide PWH with group-specific symptom management interventions based on five profiles to improve their physical and psychological well-being. Future studies should be conducted in different contexts using different symptom checklists to further validate our results.

Contemporaneous symptom networks of multidimensional symptom experiences in cancer survivors: A network analysis

BACKGROUND

Symptom networks can provide empirical evidence for the development of personalized and precise symptom management strategies. However, few studies have explored the symptom networks of multidimensional symptom experiences in cancer survivors. The objectives of this study were to generate symptom networks of multidimensional symptom experiences in cancer survivors and explore the centrality indices and density in these symptom networks METHODS: Data from 1065 cancer survivors were obtained from the Shanghai CANcer Survivor (SCANS) Report. The MD Anderson Symptom Inventory was used to assess the prevalence and severity of 13 cancer-related symptoms. We constructed contemporaneous networks with all 13 symptoms after controlling for covariates.

RESULTS

Distress (r_s  = 9.18, r_c  = 0.06), sadness (r_s  = 9.05, r_c  = 0.06), and lack of appetite (r_s  = 9.04, r_c  = 0.06) had the largest values for strength and closeness. The density of the "less than 5 years" network was significantly different from that of the "5-10 years" and "over 10 years" networks (p < 0.001). We found that while fatigue was the most severe symptom in cancer survivorship, the centrality of fatigue was lower than that of the majority of other symptoms.

CONCLUSION

Our study demonstrates the need for the assessment of centrality indices and network density as an essential component of cancer care, especially for survivors with <5 years of survivorship. Future studies are warranted to develop dynamic symptom networks and trajectories of centrality indices in longitudinal data to explore causality among symptoms and markers of interventions.

Network Analysis of Depressive Symptomatology in Underweight and Obese Adults

BACKGROUND

Depressive symptoms can affect people's quality of life and social environment. In addition, in severe situations, they can lead to suicidal behaviors.

OBJECTIVE

This study aimed to analyze the differences in depressive symptoms in underweight and obese Peruvian adults.

METHODS

A cross-sectional study was carried out based on secondary data obtained from the Instituto Nacional de Estadística e Informática (INEI), Lima, Peru. A sample of 10 053 participants was considered, of which 55.96% were women. Two Gaussian plot models were estimated and the levels of depressive symptomatology were compared between the 2 groups (adults with underweight and obese).

RESULTS

A total of 1510 (15.02%) were underweight adults and 8543 (84.98%) were obese adults. There were differences in the reporting of depressive symptoms in the underweight group; the most central items were "Depressed mood" (PH2), "Tiredness/low energy" (PH4), and "Psychomotor difficulties" (PH8).

CONCLUSION

This study provides new evidence on the dynamic relationship between depressive symptoms according to the body mass index categories (underweight and obese) assessed.

Do network centrality measures predict dental outcomes of Indigenous children over time?

BACKGROUND

Centrality measures identify items that are central to a network, which may inform potential targets for oral interventions.

AIM

We tested whether centrality measures in a cross-sectional network of mothers' baseline factors are able to predict the association with children's dental outcomes at age 5 years.

DESIGN

A network approach was applied to longitudinal data from a randomised controlled trial of dental caries prevention delivered to 448 women pregnant with an Indigenous child in South Australia. Central items were identified at baseline using three centrality measures (strength, betweenness, and closeness). Centrality values of mothers' outcomes were regressed with their predictive values to dental caries experience and dental service utilisation at child age 5 years.

RESULTS

Items of oral health self-efficacy and oral health literacy were central to mothers' baseline network. Strength at baseline explained 51% and 45% of items' predictive values to dental caries experience and dental service utilisation at child age 5 years, respectively. Adjusted and unadjusted values of node strength for the children's oral health network were highly correlated.

CONCLUSION

Strength at baseline successfully identified mothers' items with greater importance to dental caries experience and dental service utilisation at child age 5 years.

Studying Behaviour Change Mechanisms under Complexity

Understanding the mechanisms underlying the effects of behaviour change interventions is vital for accumulating valid scientific evidence, and useful to informing practice and policy-making across multiple domains. Traditional approaches to such evaluations have applied study designs and statistical models, which implicitly assume that change is linear, constant and caused by independent influences on behaviour (such as behaviour change techniques). This article illustrates limitations of these standard tools, and considers the benefits of adopting a complex adaptive systems approach to behaviour change research. It (1) outlines the complexity of behaviours and behaviour change interventions; (2) introduces readers to some key features of complex systems and how these relate to human behaviour change; and (3) provides suggestions for how researchers can better account for implications of complexity in analysing change mechanisms. We focus on three common features of complex systems (i.e., interconnectedness, non-ergodicity and non-linearity), and introduce Recurrence Analysis, a method for non-linear time series analysis which is able to quantify complex dynamics. The supplemental website provides exemplifying code and data for practical analysis applications. The complex adaptive systems approach can complement traditional investigations by opening up novel avenues for understanding and theorising about the dynamics of behaviour change.