Parsing Social Network Survey Data from Hidden Populations Using Stochastic Context-Free Grammars

Background

Human populations are structured by social networks, in which individuals tend to form relationships based on shared attributes. Certain attributes that are ambiguous, stigmatized or illegal can create a ÔhiddenÕ population, so-called because its members are difficult to identify. Many hidden populations are also at an elevated risk of exposure to infectious diseases. Consequently, public health agencies are presently adopting modern survey techniques that traverse social networks in hidden populations by soliciting individuals to recruit their peers, e.g., respondent-driven sampling (RDS). The concomitant accumulation of network-based epidemiological data, however, is rapidly outpacing the development of computational methods for analysis. Moreover, current analytical models rely on unrealistic assumptions, e.g., that the traversal of social networks can be modeled by a Markov chain rather than a branching process.

Methodology/Principal Findings

Here, we develop a new methodology based on stochastic context-free grammars (SCFGs), which are well-suited to modeling tree-like structure of the RDS recruitment process. We apply this methodology to an RDS case study of injection drug users (IDUs) in Tijuana, México, a hidden population at high risk of blood-borne and sexually-transmitted infections (i.e., HIV, hepatitis C virus, syphilis). Survey data were encoded as text strings that were parsed using our custom implementation of the inside-outside algorithm in a publicly-available software package (HyPhy), which uses either expectation maximization or direct optimization methods and permits constraints on model parameters for hypothesis testing. We identified significant latent variability in the recruitment process that violates assumptions of Markov chain-based methods for RDS analysis: firstly, IDUs tended to emulate the recruitment behavior of their own recruiter; and secondly, the recruitment of like peers (homophily) was dependent on the number of recruits.

Conclusions

SCFGs provide a rich probabilistic language that can articulate complex latent structure in survey data derived from the traversal of social networks. Such structure that has no representation in Markov chain-based models can interfere with the estimation of the composition of hidden populations if left unaccounted for, raising critical implications for the prevention and control of infectious disease epidemics.     

Socio-Economic and Health Access Determinants of Breast and Cervical Cancer Screening in Low-Income Countries: Analysis of the World Health Survey

Background

Breast and Cervical cancer are the two most common cancers among women in developing countries. Regular screening is the most effective way of ensuring that these cancers are detected at early stages; however few studies have assessed factors that predict cancer screening in developing countries.

Purpose

To assess the influence of household socio-economic status (SES), healthcare access and country level characteristics on breast and cervical cancer screening among women in developing countries.

Methods

Women ages 18–69 years (cervical cancer screening) and 40–69 years (breast cancer screening) from 15 developing countries who participated in the 2003 World Health Survey provided data for this study. Household SES and healthcare access was assessed based on self-reported survey responses. SAS survey procedures (SAS, Version 9.2) were used to assess determinants of breast and cervical cancer screening in separate models.

Results

4.1% of women ages 18–69 years had received cervical cancer screening in the past three years, while only 2.2% of women ages 40–69 years had received breast cancer screening in the past 5 years in developing countries. Cancer screening rates varied by country; cervical cancer screening ranged from 1.1% in Bangladesh to 57.6% in Congo and breast cancer screening ranged from 0% in Mali to 26% in Congo. Significant determinants of cancer screening were household SES, rural residence, country health expenditure (as a percent of GDP) as well as healthcare access.

Discussion

A lot more needs to be done to improve screening rates for breast and cervical cancer in developing countries, such as increasing health expenditure (especially in rural areas), applying the increased funds towards the provision of more, better educated health providers as well as improved infrastructure.     

The Relationship between Livestock Ownership and Child Stunting in Three Countries in Eastern Africa Using National Survey Data

Livestock ownership has the potential to improve child nutrition through various mechanisms, although direct evaluations of household livestock and child stunting status are uncommon. We conducted an analysis of Demographic and Health Survey (DHS) datasets from Ethiopia (2011), Kenya (2008–2009), and Uganda (2010) among rural children under 5 years of age to compare stunting status across levels of livestock ownership. We classified livestock ownership by summing reported household numbers of goats, sheep, cattle and chickens, as well as calculating a weighted score to combine multiple species. The primary association was assessed separately by country using a log-binomial model adjusted for wealth and region, which was then stratified by child diarrheal illness, animal-source foods intake, sub-region, and wealth index. This analysis included n = 8079 children from Ethiopia, n = 3903 children from Kenya, and n = 1645 from Uganda. A ten-fold increase in household livestock ownership had significant association with lower stunting prevalence in Ethiopia (Prevalence Ratio [PR] 0.95, 95% CI 0.92–0.98) and Uganda (PR 0.87, 95% CI 0.79–0.97), but not Kenya (PR 1.01, 95% CI 0.96–1.07). The weighted livestock score was only marginally associated with stunting status. The findings varied slightly by region, but not by wealth, diarrheal disease, or animal-source food intake. This analysis suggested a slightly beneficial effect of household livestock ownership on child stunting prevalence. The small effect size observed may be related to limitations of the DHS dataset or the potentially complicated relationship between malnutrition and livestock ownership, including livestock health and productivity.     

Identifying Signatures of Natural Selection in Tibetan and Andean Populations Using Dense Genome Scan Data

High-altitude hypoxia (reduced inspired oxygen tension due to decreased barometric pressure) exerts severe physiological stress on the human body. Two high-altitude regions where humans have lived for millennia are the Andean Altiplano and the Tibetan Plateau. Populations living in these regions exhibit unique circulatory, respiratory, and hematological adaptations to life at high altitude. Although these responses have been well characterized physiologically, their underlying genetic basis remains unknown. We performed a genome scan to identify genes showing evidence of adaptation to hypoxia. We looked across each chromosome to identify genomic regions with previously unknown function with respect to altitude phenotypes. In addition, groups of genes functioning in oxygen metabolism and sensing were examined to test the hypothesis that particular pathways have been involved in genetic adaptation to altitude. Applying four population genetic statistics commonly used for detecting signatures of natural selection, we identified selection-nominated candidate genes and gene regions in these two populations (Andeans and Tibetans) separately. The Tibetan and Andean patterns of genetic adaptation are largely distinct from one another, with both populations showing evidence of positive natural selection in different genes or gene regions. Interestingly, one gene previously known to be important in cellular oxygen sensing, EGLN1 (also known as PHD2), shows evidence of positive selection in both Tibetans and Andeans. However, the pattern of variation for this gene differs between the two populations. Our results indicate that several key HIF-regulatory and targeted genes are responsible for adaptation to high altitude in Andeans and Tibetans, and several different chromosomal regions are implicated in the putative response to selection. These data suggest a genetic role in high-altitude adaption and provide a basis for future genotype/phenotype association studies necessary to confirm the role of selection-nominated candidate genes and gene regions in adaptation to altitude.     

Social Determinants of Health in Environmental Justice Communities: Examining Cumulative Risk in Terms of Environmental Exposures and Social Determinants of Health

Residents of environmental justice (EJ) communities may bear a disproportionate burden of environmental health risk, and often face additional burdens from social determinants of health. Accounting for cumulative risk should include measures of risk from both environmental sources and social determinants. This study sought to better understand cumulative health risk from both social and environmental sources in a disadvantaged community in Texas. Key outcomes were determining what data are currently available for this assessment, clarifying data needs, identifying data gaps, and considering how those gaps could be filled. Analyses suggested that the traditionally defined EJ community in Port Arthur may have a lower environmental risk from air toxics than the rest of the City of Port Arthur (although the entire city has a higher risk than the average for the state), but may have a larger burden from social determinants of health. However, the results should be interpreted in light of the availability of data, the definitions of community boundaries, and the areal unit utilized. Continued focus on environmental justice communities and the cumulative risks faced by their residents is critical to protecting these residents and, ultimately, moving toward a more equitable distribution and acceptable level of risk throughout society.     

BCG-Mediated Bladder Cancer Immunotherapy: Identifying Determinants of Treatment Response Using a Calibrated Mathematical Model

Intravesical Bacillus Calmette Guérin (BCG) immunotherapy is considered the standard of care for treatment of non-muscle invasive bladder cancer; however the treatment parameters were established empirically. In order to evaluate potential optimization of clinical parameters of BCG induction therapy, we constructed and queried a new mathematical model. Specifically, we assessed the impact of (1) duration between resection and the first instillation; (2) BCG dose; (3) indwelling time; and (4) treatment interval of induction therapy – using cure rate as the primary endpoint. Based on available clinical and in vitro experimental data, we constructed and parameterized a stochastic mathematical model describing the interactions between BCG, the immune system, the bladder mucosa and tumor cells. The primary endpoint of the model was the probability of tumor extinction following BCG induction therapy in patients with high risk for tumor recurrence. We theoretically demonstrate that extending the duration between the resection and the first BCG instillation negatively influences treatment outcome. Simulations of higher BCG doses and longer indwelling times both improved the probability of tumor extinction. A remarkable finding was that an inter-instillation interval two times longer than the seven-day interval used in the current standard of care would substantially improve treatment outcome. We provide insight into relevant clinical questions using a novel mathematical model of BCG immunotherapy. Our model predicts an altered regimen that may decrease side effects of treatment while improving response to therapy.     

Functional Dentition in Brazilian Adults: An Investigation of Social Determinants of Health (SDH) Using a Multilevel Approach

Objectives

Estimate the prevalence of functional dentition among Brazilian adults using four different definitions and identify associated factors.

Methods

A cross-sectional study was conducted involving 9564 Brazilian adults aged 35–44 years who participated in the 2010 National Oral Health Survey. Data collection involved oral examinations and the administration of questionnaires. The following definitions were used: 1—WHO Functional Dentition (FDWHO: ≥ 20 teeth present); 2—well-distributed teeth (WDT: ≥ 10 teeth in each arch); 3 –Functional dentition classified by esthetics and occlusion (FD_Class5: dentitions that sequentially exhibit at least one tooth in each arch, at least 10 teeth in each arch, all maxillary and mandibular anterior teeth, three or four premolar posterior occluding pairs [POPs], and at least one molar POP bilaterally); 4—Functional dentition classified by esthetics, occlusion and periodontal status (FD_Class6: corresponds to FD_Class5 with the addition of periodontal status of all sextants in the oral cavity with, at most, shallow pockets and/or clinical attachment level of 5 mm (CPI ≤ 3 and/or CAL ≤ 1). The independent variables were individual factors (gender, self-declared skin color, schooling, monthly household income, age group, self-rated treatment need, dental pain, dental appointment in the previous 12 months and dental services) and contextual factors (Municipal Human Development Index [MHDI]), Gini coefficient, fluoridated water supply and oral health coverage). Multilevel mixed-effect Poisson regression analyses were performed.

Results

The prevalence of functional dentition based on the FDWHO, WDT, FD_Class5 and FD_Class6 definitions was 77.9%, 72.9%, 42.6% and 40.3%, respectively. Adults with ≥12 years of schooling and monthly household income from US$ 853 to 2557 had higher prevalence rates of FDWHO (PR: 1.41 and 1.10, respectively), WDT (PR: 1.58 and 1.14, respectively), FD_Class5 (PR: 2.03 and 1.27, respectively) and FD_Class6 (PR: 2.15 and 1.35, respectively). These values in the final models were adjusted for gender, self-declared skin color (FD_Class5), age group, self-rated treatment need (FDWHO, FD_Class5 and FD_Class6), dental appointment in the previous 12 months (FDWHO and WDT), dental services (FDWHO and WDT) and contextual factors. A very high MHDI and presence of fluoridated water supply were associated with higher prevalence rates of the four outcomes.

Conclusions

The incorporation of the criteria of new definitions of functional dentition led to a lower prevalence rate among Brazilian adults. Striking individual and contextual inequalities were identified with regard to the four definitions analyzed, which need to be addressed through inter-sector efforts.     

Social Support and Health in Diabetes Patients: An Observational Study in Six European Countries in an Era of Austerity

Introduction

Support from individual social networks, community organizations and neighborhoods is associated with better self-management and health outcomes. This international study examined the relative impact of different types of support on health and health-related behaviors in patients with type 2 diabetes.

Methods

Observational study (using interviews and questionnaires) in a sample of 1,692 type 2 diabetes patients with 5,433 connections from Bulgaria, Greece, Netherlands, Norway, Spain, and the United Kingdom. Outcomes were patient-reported health status (SF-12), physical exercise (RAPA), diet and smoking (SDCSCA). Random coefficient regression models were used to examine linkages with individual networks, community organizations, and neighborhood type (deprived rural, deprived urban, or affluent urban).

Results

Patients had a median of 3 support connections and 34.6% participated in community organizations. Controlled for patients’ age, sex, education, income and comorbidities, large emotional support networks were associated with decrease of non-smoking (OR = 0.87). Large practical support networks were associated with worse physical and mental health (B = -0.46 and -0.27 respectively) and less physical activity (OR = 0.90). Participation in community organizations was associated with better physical and mental health (B = 1.39 and 1.22, respectively) and, in patients with low income, with more physical activity (OR = 1.53).

Discussion

Participation in community organizations was most consistently related to better health status. Many diabetes patients have individual support networks, but this study did not provide evidence to increase their size as a public health strategy. The consistent association between participation in community organizations and health status provides a clear target for interventions and policies.     

Identifying and Targeting Mortality Disparities: A Framework for Sub-Saharan Africa Using Adult Mortality Data from South Africa

Background

Health inequities in developing countries are difficult to eradicate because of limited resources. The neglect of adult mortality in Sub-Saharan Africa (SSA) is a particular concern. Advances in data availability, software and analytic methods have created opportunities to address this challenge and tailor interventions to small areas. This study demonstrates how a generic framework can be applied to guide policy interventions to reduce adult mortality in high risk areas. The framework, therefore, incorporates the spatial clustering of adult mortality, estimates the impact of a range of determinants and quantifies the impact of their removal to ensure optimal returns on scarce resources.

Methods

Data from a national cross-sectional survey in 2007 were used to illustrate the use of the generic framework for SSA and elsewhere. Adult mortality proportions were analyzed at four administrative levels and spatial analyses were used to identify areas with significant excess mortality. An ecological approach was then used to assess the relationship between mortality “hotspots” and various determinants. Population attributable fractions were calculated to quantify the reduction in mortality as a result of targeted removal of high-impact determinants.

Results

Overall adult mortality rate was 145 per 10,000. Spatial disaggregation identified a highly non-random pattern and 67 significant high risk local municipalities were identified. The most prominent determinants of adult mortality included HIV antenatal sero-prevalence, low SES and lack of formal marital union status. The removal of the most attributable factors, based on local area prevalence, suggest that overall adult mortality could be potentially reduced by ∼90 deaths per 10,000.

Conclusions

The innovative use of secondary data and advanced epidemiological techniques can be combined in a generic framework to identify and map mortality to the lowest administration level. The identification of high risk mortality determinants allows health authorities to tailor interventions at local level. This approach can be replicated elsewhere.     

Factors Associated with Untreated Diabetes: Analysis of Data from 20,496 Participants in the Japanese National Health and Nutrition Survey

Objective

We aimed to examine factors associated with untreated diabetes in a nationally representative sample of the Japanese population.

Research Design and Methods

We pooled data from the Japanese National Health and Nutrition Survey from 2005 to 2009 (n = 20,496). Individuals aged 20 years and older were included in the analysis. We classified participants as having diabetes if they had HbA1c levels ≥6.5% (≥48 mmol/mol). People with diabetes who self-reported that they were not currently receiving diabetic treatment were considered to be untreated. We conducted a multinomial logistic regression analysis to determine factors associated with untreated diabetes relative to non-diabetic individuals.

Results

Of 20,496 participants who were included in the analysis, untreated diabetes was present in 748 (3.6%). Among participants with untreated diabetes, 48.3% were previously diagnosed with diabetes, and 46.5% had HbA1c levels ≥7.0% (≥53 mmol/mol). Participants with untreated diabetes were significantly more likely than non-diabetic participants to be male, older, and currently smoking, have lower HDL cholesterol levels and higher BMI, non-HDL cholesterol levels, and systolic blood pressure.

Conclusions

A substantial proportion of people in Japan with untreated diabetes have poor glycemic control. Targeting relevant factors for untreated diabetes in screening programs may be effective to enhance the treatment and control of diabetes.     

Changes in relative fitness and frailty across the adult lifespan: evidence from the Canadian National Population Health Survey

Background

The prevalence of frailty increases with age in older adults, but frailty is largely unreported for younger adults, where its associated risk is less clear. Furthermore, less is known about how frailty changes over time among younger adults. We estimated the prevalence and outcomes of frailty, in relation to accumulation of deficits, across the adult lifespan.

Methods

We analyzed data for community-dwelling respondents (age 15–102 years at baseline) to the longitudinal component of the National Population Health Survey, with seven two-year cycles, beginning 1994–1995. The outcomes were death, use of health services and change in health status, measured in terms of a Frailty Index constructed from 42 self-reported health variables.

Results

The sample consisted of 14 713 respondents (54.2% women). Vital status was known for more than 99% of the respondents. The prevalence of frailty increased with age, from 2.0% (95% confidence interval [CI] 1.7%–2.4%) among those younger than 30 years to 22.4% (95% CI 19.0%–25.8%) for those older than age 65, including 43.7% (95% CI 37.1%–50.8%) for those 85 and older. At all ages, the 160-month mortality rate was lower among relatively fit people than among those who were frail (e.g., 2% v. 16% at age 40; 42% v. 83% at age 75 or older). These relatively fit people tended to remain relatively fit over time. Relative to all other groups, a greater proportion of the most frail people used health services at baseline (28.3%, 95% CI 21.5%–35.5%) and at each follow-up cycle (26.7%, 95% CI 15.4%–28.0%).

Interpretation

Deficits accumulated with age across the adult spectrum. At all ages, a higher Frailty Index was associated with higher mortality and greater use of health care services. At younger ages, recovery to the relatively fittest state was common, but the chance of complete recovery declined with age.On average, health declines with age. Even so, at any given age the health status across a group of people varies. Variability in health status and in the risk for adverse outcomes for people of the same age is referred to as “frailty,” which typically has been studied among older adults.^1,2 Although frailty can be operationalized in different ways, in general, people who report having no health problems are more likely to be fit than people who report having many problems. Unsurprisingly, the chance of adverse outcomes — death, admission to a long-term care institution or to hospital, or worsening of health status — increases with the number of problems that the individual has.^3,4The antecedents of frailty appear to arise some time before old age,^5–9 although how frailty emerges as people age, whether it carries the same risk at all ages and the extent to which it fluctuates are less clear.^9,10 In the study reported here, we evaluated changes in relative fitness and frailty across the adult lifespan. Our objectives were to investigate the effect of age on the prevalence of relative fitness and frailty, the characteristics of people who were relatively fit in comparison with those who were frail across the adult lifespan, the effects of fitness and frailty on mortality in relation to age and sex, and the characteristics of people who maintained the highest levels of fitness across a decade relative to those who at any point reported any decline.     

Metaproteomic Survey of Six Aquatic Habitats: Discovering the Identities of Microbial Populations Active in Biogeochemical Cycling

Our goal is to strengthen the foundations of metaproteomics as a microbial community analysis tool that links the functional identity of actively expressed gene products with host phylogeny. We used shotgun metaproteomics to survey waters in six disparate aquatic habitats (Cayuga Lake, NY; Oneida Lake, NY; Gulf of Maine; Chesapeake Bay, MD; Gulf of Mexico; and the South Pacific). Peptide pools prepared from filter-gathered microbial biomass, analyzed by nano-liquid chromatography–mass spectrometry (MS/MS) generating 9,693?±?1,073 mass spectra identified 326?±?107 bacterial proteins per sample. Distribution of proteobacterial (Alpha and Beta) and cyanobacterial (Prochlorococcus and Synechococcus spp.) protein hosts across all six samples was consistent with the previously published biogeography for these microorganisms. Marine samples were enriched in transport proteins (TRAP-type for dicarboxylates and ATP binding cassette (ABC)-type for amino acids and carbohydrates) compared with the freshwater samples. We were able to match in situ expression of many key proteins catalyzing C-, N-, and S-cycle processes with their bacterial hosts across all six habitats. Pelagibacter was identified as the host of ABC-type sugar-, organic polyanion-, and glycine betaine-transport proteins; this extends previously published studies of Pelagibacter's in situ biogeochemical role in marine C- and N-metabolism. Proteins matched to Ruegeria confirmed these organism's role in marine waters oxidizing both carbon monoxide and sulfide. By documenting both processes expressed in situ and the identity of host cells, metaproteomics tested several existing hypotheses about ecophysiological processes and provided fodder for new ones.     

Shortest-Path Network Analysis Is a Useful Approach toward Identifying Genetic Determinants of Longevity

Background

Identification of genes that modulate longevity is a major focus of aging-related research and an area of intense public interest. In addition to facilitating an improved understanding of the basic mechanisms of aging, such genes represent potential targets for therapeutic intervention in multiple age-associated diseases, including cancer, heart disease, diabetes, and neurodegenerative disorders. To date, however, targeted efforts at identifying longevity-associated genes have been limited by a lack of predictive power, and useful algorithms for candidate gene-identification have also been lacking.

Methodology/Principal Findings

We have utilized a shortest-path network analysis to identify novel genes that modulate longevity in Saccharomyces cerevisiae. Based on a set of previously reported genes associated with increased life span, we applied a shortest-path network algorithm to a pre-existing protein–protein interaction dataset in order to construct a shortest-path longevity network. To validate this network, the replicative aging potential of 88 single-gene deletion strains corresponding to predicted components of the shortest-path longevity network was determined. Here we report that the single-gene deletion strains identified by our shortest-path longevity analysis are significantly enriched for mutations conferring either increased or decreased replicative life span, relative to a randomly selected set of 564 single-gene deletion strains or to the current data set available for the entire haploid deletion collection. Further, we report the identification of previously unknown longevity genes, several of which function in a conserved longevity pathway believed to mediate life span extension in response to dietary restriction.

Conclusions/Significance

This work demonstrates that shortest-path network analysis is a useful approach toward identifying genetic determinants of longevity and represents the first application of network analysis of aging to be extensively validated in a biological system. The novel longevity genes identified in this study are likely to yield further insight into the molecular mechanisms of aging and age-associated disease.     

Genetic Determinants of Facial Clefting: Analysis of 357 Candidate Genes Using Two National Cleft Studies from Scandinavia

Background

Facial clefts are common birth defects with a strong genetic component. To identify fetal genetic risk factors for clefting, 1536 SNPs in 357 candidate genes were genotyped in two population-based samples from Scandinavia (Norway: 562 case-parent and 592 control-parent triads; Denmark: 235 case-parent triads).

Methodology/Principal Findings

We used two complementary statistical methods, TRIMM and HAPLIN, to look for associations across these two national samples. TRIMM tests for association in each gene by using multi-SNP genotypes from case-parent triads directly without the need to infer haplotypes. HAPLIN on the other hand estimates the full haplotype distribution over a set of SNPs and estimates relative risks associated with each haplotype. For isolated cleft lip with or without cleft palate (I-CL/P), TRIMM and HAPLIN both identified significant associations with IRF6 and ADH1C in both populations, but only HAPLIN found an association with FGF12. For isolated cleft palate (I-CP), TRIMM found associations with ALX3, MKX, and PDGFC in both populations, but only the association with PDGFC was identified by HAPLIN. In addition, HAPLIN identified an association with ETV5 that was not detected by TRIMM.

Conclusion/Significance

Strong associations with seven genes were replicated in the Scandinavian samples and our approach effectively replicated the strongest previously known association in clefting—with IRF6. Based on two national cleft cohorts of similar ancestry, two robust statistical methods and a large panel of SNPs in the most promising cleft candidate genes to date, this study identified a previously unknown association with clefting for ADH1C and provides additional candidates and analytic approaches to advance the field.     

Health Insurance Status as a Barrier to Ideal Cardiovascular Health for U.S. Adults: Data from the National Health and Nutrition Examination Survey (NHANES)

Background

Little is known about the association between cardiovascular (CV) health and health insurance status. We hypothesized that U.S. adults without health insurance coverage would have a lower likelihood of ideal cardiovascular health.

Methods and Results

Using National Health and Nutrition Examination Survey (NHANES) data from 2007–2010, we examined the relationship between health insurance status and ideal CV health in U.S. adults aged ≥19 years and <65 (N = 3304). Ideal CV health was defined by the American Heart Association (AHA) as the absence of clinically manifested CV disease and the simultaneous presence of 6–7 “ideal” CV health factors and behaviors. Logistic regression modeling was used to determine the relationship between health insurance status and the odds of ideal CV health. Of the U.S. adult population, 5.4% attained ideal CV health, and 23.5% were without health insurance coverage. Those without health insurance coverage were more likely to be young (p<0.0001), male (p<0.0001), non-white (p<0.0001), with less than a high school degree (p<0.0001), have a poverty-to-income ratio less than 1 (p<0.0001) and unemployed (p<0.0001) compared to those with coverage. Lack of health insurance coverage was associated with a lower likelihood of ideal CV health; however, this relationship was attenuated by socioeconomic status.

Conclusions

U.S. adults without health insurance coverage are less likely to have ideal CV health. Population-based strategies and interventions directed at the community-level may be one way to improve overall CV health and reach this at-risk group.     

Multi-Scale Compositionality: Identifying the Compositional Structures of Social Dynamics Using Deep Learning

Objective

Social media exhibit rich yet distinct temporal dynamics which cover a wide range of different scales. In order to study this complex dynamics, two fundamental questions revolve around (1) the signatures of social dynamics at different time scales, and (2) the way in which these signatures interact and form higher-level meanings.

Method

In this paper, we propose the Recursive Convolutional Bayesian Model (RCBM) to address both of these fundamental questions. The key idea behind our approach consists of constructing a deep-learning framework using specialized convolution operators that are designed to exploit the inherent heterogeneity of social dynamics. RCBM’s runtime and convergence properties are guaranteed by formal analyses.

Results

Experimental results show that the proposed method outperforms the state-of-the-art approaches both in terms of solution quality and computational efficiency. Indeed, by applying the proposed method on two social network datasets, Twitter and Yelp, we are able to identify the compositional structures that can accurately characterize the complex social dynamics from these two social media. We further show that identifying these patterns can enable new applications such as anomaly detection and improved social dynamics forecasting. Finally, our analysis offers new insights on understanding and engineering social media dynamics, with direct applications to opinion spreading and online content promotion.     

Explorative Analysis of Wuhan Intra-Urban Human Mobility Using Social Media Check-In Data

Social media check-in data as a geo-tagged information source have been used for revealing spatio-temporal patterns in the field of social and urban study, such as human behavior or public issues. This paper investigates a case study and presents a new method of representing the mobility of people within a city from check-in data. By dividing the data in a temporal sequence, this study examines the overall mobility in the case study city through the gradient/difference of population density with a series of time after computing the population density from the check-in data using an incorporated Thiessen polygon method. By classifying check-in data with their geo-tags into several groups according to travel purposes, and partitioning the data according to administrative district boundaries, various moving patterns for those travel purposes in those administrative districts are identified by scrutinizing a series of spatial geometries of a weighted standard deviational ellipse (WSDE). Through deep analyses of those data by the adopted approaches, the general pattern of mobility in the case city, such as people moving to the central urban area from the suburb from 4 am to 8 am, is ascertained, and different characteristics of movement in those districts are also depicted. Furthermore, it can tell that in which district less movement is likely for a certain purpose (e.g., for dinner or entertainment). This study has demonstrated the availability of the proposed methodology and check-in data for investigating intra-urban human mobility.     

Determinants of Salt-Restriction-Spoon Using Behavior in China: Application of the Health Belief Model

Background

The two-gram salt-restriction-spoons, which can be used to reduce the salt intake of people, had been handed out for free by the Chinese government to the citizens several years ago, but only a small fraction of residents use such a spoon currently. Since no studies have been conducted to investigate relevant influencing factors, this study was designed to explore the determinants of salt-restriction-spoon using behavior (SRB) in China.

Methods

This cross-sectional study was conducted in Beijing, China. Altogether 269 rural residents and 244 urban residents aged over 18 were selected by convenience sampling method in 2012. Variables measured in a questionnaire designed according to the Health Belief Model (HBM) included socio-demographics, perceived susceptibility, perceived severity, perceived benefits, perceived objective barriers, perceived subjective barriers, self-efficacy, knowledge of hypertension, cues to action, and SRB. Answers to the questionnaire were obtained from all the participants, and 24-hour urine samples were collected to determine the 24-hour urinary sodium excretion (24HUNa). Path analyses were used to explore the determinants of SRB.

Results

Approximately 22.7% and 45.3% of residents used a salt-restriction-spoon everyday in the rural and urban areas, respectively. The average 24HUNa was 211.19±98.39 mmol for rural residents and 109.22±58.18 mmol for urban residents. Path analyses shown that perceived objective barriers, perceived benefits, perceived severity, knowledge and age were related to SRB and 24HUNa for both rural and urban participants, among which perceived objective barrier (β =  − 0.442 and β =  − 0.543, respectively) was the most important determinant.

Conclusion

Improvement of the current salt-restriction-spoon and education on the right usage of the salt-restriction-spoon, the severity of hypertension, and the benefit of salt reduction are necessary, especially among those who are relatively young but at risk of hypertension, those who have lower education levels, and those who live in the rural areas.     

Reference Values of Skeletal Muscle Mass for Korean Children and Adolescents Using Data from the Korean National Health and Nutrition Examination Survey 2009-2011

BackgroundSkeletal muscle mass (SMM) plays a crucial role in systemic glucose metabolism.ObjectiveTo obtain reference data on absolute and relative values of SMM for Korean children and adolescents.MethodsCross-sectional results from 1919 children and adolescents (1024 boys) aged 10–18 years that underwent dual-energy X-ray absorptiometry (DXA) during the Korean National Health and Nutrition Examination Survey 2009–2011 were analyzed. SMMs were evaluated as follows; absolute SMM (appendicular skeletal muscle mass [ASM]) and relative SMMs, namely, height-adjusted skeletal muscle index (SMI; ASM/height²), %SMM (ASM/weight x 100), and skeletal muscle-to-body fat ratio (MFR; ASM/body fat mass).ResultsPercentile curves illustrated the developmental patterns of the SMMs of Korean children and adolescents. ASM and SMI increased with age in both genders, and increased from age 10 throughout adolescence in boys, whereas in girls, they increased until age 13 and then stabilized. In boys, %SMM and MFR were highest at age 15 and then slowly stabilized or decreased, whereas in girls, they peaked at age 10 to 11 and then decreased through adolescence. Cut-off values for low MFR were identified and a significant association was found between a low MFR and high risk of metabolic syndrome. However, this association was found to be dependent on gender and the level of BMI.ConclusionThis study provides reference values of absolute and relative SMM for Korean children and adolescents. Detailed body composition analyses including skeletal muscle and fat mass might provide improved measures of metabolic risk.