Embedded restricted Boltzmann machines for fusion of mixed data types and applications in social measurements analysis

Analysis and fusion of social measurements is important to understand what shapes the public’s opinion and the sustainability of the global development. However, modeling data collected from social responses is challenging as the data is typically complex and heterogeneous, which might take the form of stated facts, subjective assessment, choices, preferences or any combination thereof. Model-wise, these responses are a mixture of data types including binary, categorical, multicategorical, continuous, ordinal, count and rank data. The challenge is therefore to effectively handle mixed data in the a unified fusion framework in order to perform inference and analysis. To that end, this paper introduces eRBM (Embedded Restricted Boltzmann Machine) – a probabilistic latent variable model that can represent mixed data using a layer of hidden variables transparent across different types of data. The proposed model can comfortably support largescale data analysis tasks, including distribution modelling, data completion, prediction and visualisation. We demonstrate these versatile features on several moderate and large-scale publicly available social survey datasets.

Predicting steep escalations in alcohol use over the teenage years: age-related variations in key social influences

Aims
This study examined how family, peer and school factors are related to different trajectories of adolescent alcohol use at key developmental periods.

Design
Latent class growth analysis was used to identify trajectories based on five waves of data (from grade 6, age 12 to grade 11, age 17), with predictors at grades 5, 7 and 9 included as covariates.

Setting
Adolescents completed surveys during school hours.

Participants
A total of 808 students in Victoria, Australia.

Measurements
Alcohol use trajectories were based on self-reports of 30-day frequency of alcohol use. Predictors included sibling alcohol use, attachment to parents, parental supervision, parental attitudes favourable to adolescent alcohol use, peer alcohol use and school commitment.

Findings
A total of 8.2% showed steep escalation in alcohol use. Relative to non-users, steep escalators were predicted by age-specific effects for low school commitment at grade 7 (P = 0.031) and parental attitudes at grade 5 (P = 0.003), and age-generalized effects for sibling alcohol use (Ps = 0.001, 0.012, 0.033 at grades 5, 7 and 9, respectively) and peer alcohol use (Ps = 0.041, < 0.001, < 0.001 at grades 5, 7 and 9, respectively). Poor parental supervision was associated with steep escalators at grade 9 (P < 0.001) but not the other grades. Attachment to parents was unrelated to alcohol trajectories.

Conclusions
Parental disapproval of alcohol use before transition to high school, low school commitment at transition to high school, and sibling and peer alcohol use during adolescence are associated with a higher risk of steep escalations in alcohol use.

Different patterns of Australian adults' knowledge of foods and nutrients related to metabolic disease risk

A nationwide survey of 2022 consumers was conducted in Australia in late 2011. A short list of questions about knowledge of the nutrient composition of common foods was administered along with questions about the respondents’ food attitudes, demographics, school education and dieting practices. Overall, the results showed that nutrition knowledge was relatively high. Latent class analysis showed two groups of consumers with ‘high’ and ‘low’ knowledge of nutrition. Higher knowledge was positively associated with age, female sex, university education, experience of home economics or health education at school, having a chronic disease, and attitudes to food issues, and negatively with type 1 diabetes or the use of diabetes-control diets. The implications of the findings for nutrition communication are discussed.

When social anxiety co-occurs with substance use: does an impulsive social anxiety subtype explain this unexpected relationship?

Although most conceptualizations of social anxiety emphasise that socially anxious individuals are overtly shy, and utilise avoidant behavioural strategies (e.g., risk-aversion, passivity, and submissiveness), there is tentative support for the existence of an approach-motivated subtype, characterised by risk taking and a greater propensity for substance misuse. It is likely that this subtype may help explain the reported co-occurrence of substance misuse and social anxiety. The current study sought to test via latent class analysis whether an approach-motivated social anxiety subtype could be identified within a community sample. A self-report questionnaire was completed by 351 participants (age: 18-74 years). Two distinct social anxiety subgroups were identified: one characterised by prototypical SAD symptomatology (i.e., behavioural inhibition and risk-avoidance), the second by elevated levels of rash impulsiveness, reward sensitivity, risk-taking and co-occurring substance use problems. The current findings provides support for the existence of a distinct approach-motivated social anxiety subtype and indicates that impulsivity may be critical to understanding the comorbid substance use symptomatology of these individuals.

Australian adults’ knowledge of Australian agriculture

Purpose – Agriculture is a major generator of wealth and employment in Australia. However, it faces a range of economic and environmental challenges which require substantial community support. The purpose of this paper is to examine Australian adults’ Australian knowledge of, and attitudes towards, Australian agriculture. Design/methodology/approach – Online questionnaire survey of 1,026 adults conducted nationwide during August 2012. Findings – Most respondents had little knowledge of even the basic aspects of the industry but they approved of farmers’ performance of their roles. Latent class analysis showed that there are two groups of consumers with low and lower levels of knowledge. The respondents’ age, rural residence and universalist values were positive predictors of agricultural knowledge. Research limitations/implications – This was a cross-sectional, quota-based survey which examined only some aspects of agriculture. However, the findings suggest that more communication with the general public about the industry is required in order to build on the positive sentiment that exists within the community. Practical implications – More education about agriculture in schools and higher education is indicated. Social implications – The poor state of knowledge of agriculture threatens the social contract upon which agricultural communities depend for survival. Originality/value – The study highlights the poor state of general knowledge about agriculture in Australia. The findings could be used as a baseline against which the efficacy of future education programmes could be assessed.

Typology of alcohol consumers in five Australian nighttime entertainment districts

INTRODUCTION AND AIMS: Understanding how types of alcohol consumers differ is important for public policy targeted at reducing adverse events. The aims of the present study were to identify typologies of alcohol consumers in Australian nighttime entertainment districts based on risk factors for harm and to examine variation between the identified groups in drinking setting and harms. DESIGN AND METHODS: Street-intercept surveys were conducted with 5556 alcohol consumers in and around licensed venues in five Australian cities between November 2011 and June 2012. Latent class analysis identified groups based on age and sex, and blood alcohol concentration, pre-drinking, energy drink use and illicit drug use during that night. RESULTS: Four classes were identified: general patron group (33%), young pre-drinker group (27%), intoxicated male pre-drinker group (31%) and intoxicated illicit drug male group (9%). The proportion of the general patron group interviewed decreased over the night, while the other groups increased (particularly in regional cities). As compared with the general patron group, the remaining three groups reported increased odds of being involved in aggression and any alcohol-related injuries in the past 3 months, with highest rates of harm amongst the intoxicated illicit drug male group. DISCUSSION AND CONCLUSIONS: Alcohol consumers in nighttime entertainment districts are not a homogeneous group. One-third have a low likelihood of risky consumption practices; however, representation of this consumer class diminishes throughout the night. Elevated harms amongst groups characterised by certain risk factors (e.g. pre-drinking and illicit drug use) emphasise the importance of addressing these behaviours in public policy.

Developmental trajectories of adolescent cannabis use and their relationship to young adult social and behavioural adjustment: a longitudinal study of Australian youth

This study aimed to identify distinct developmental trajectories (sub-groups of individuals who showed similar longitudinal patterns) of cannabis use among Australian adolescents, and to examine associations between trajectory group membership and measures of social and behavioural adjustment in young adulthood. Participants (n=852, 53% female) were part of the International Youth Development Study. Latent class growth analysis was used to identify distinct trajectories of cannabis use frequency from average ages 12 to 19, across 6 waves of data. Logistic regression analyses and analyses of covariance were used to examine relationships between trajectory group membership and young adult (average age: 21) adjustment, controlling for a range of covariates. Three trajectories were identified: abstainers (62%), early onset users (11%), and late onset occasional users (27%). The early onset users showed a higher frequency of antisocial behaviour, violence, cannabis use, cannabis-related harms, cigarette use, and alcohol harms, compared to the abstinent group in young adulthood. The late onset occasional users reported a higher frequency of cannabis use, cannabis-related harms, illicit drug use, and alcohol harms, compared to the abstinent group in young adulthood. There were no differences between the trajectory groups on measures of employment, school completion, post-secondary education, income, depression/anxiety, or alcohol use problems. In conclusion, early onset of cannabis use, even at relatively low frequency during adolescence, is associated with poorer adjustment in young adulthood. Prevention and intervention efforts to delay or prevent uptake of cannabis use should be particularly focussed on early adolescence prior to age 12.

Estimation of the diagnostic accuracy of the glyco-C and US9 gene-based polymerase chain reaction technique for the detection of bovine Herpesvirus type 5 DNA in decomposed brain suspension from a slaughter house using Bayesian analysis, Brazil

Fundação de Amparo à Pesquisa do Estado de São Paulo (FAPESP)

The course of chronic and recurrent low back pain in the general population

Using latent class analysis (LCA), a previous study on patients attending primary care identified four courses of low back pain (LBP) over the subsequent 6 months. To date, no studies have used longitudinal pain recordings to examine the "natural" course of recurrent and chronic LBP in a population-based sample of individuals. This study examines the course of LBP in the general population and elaborates on the stability and criterion-related validity of the clusters derived. A random sample of 400 individuals reporting LBP in a population-based study was asked to complete a comprehensive questionnaire at the start and end of the year's survey, and 52 weekly pain diaries in between. The latter were analyzed using LCA. 305 individuals returned more than 50% of the diaries. Four clusters were identified (severe persistent, moderate persistent, mild persistent, and fluctuating). The clusters differed significantly with regards to pain and disability. Assessment of cluster stability showed that a considerable proportion of patients in the "fluctuating" group changed their classification over time. Three of the four clusters describing the typical course of pain matched the clusters described previously for patients in primary care. Due to the population-based design, this study achieves, for the first time, a close insight into the "natural" course of chronic and recurrent low back pain, including individuals that did not necessarily visit the general practitioner. The findings will help to understand better the nature of this pain in the general population.

Phenotypes of childhood asthma: are they real?

It has been suggested that there are several distinct phenotypes of childhood asthma or childhood wheezing. Here, we review the research relating to these phenotypes, with a focus on the methods used to define and validate them. Childhood wheezing disorders manifest themselves in a range of observable (phenotypic) features such as lung function, bronchial responsiveness, atopy and a highly variable time course (prognosis). The underlying causes are not sufficiently understood to define disease entities based on aetiology. Nevertheless, there is a need for a classification that would (i) facilitate research into aetiology and pathophysiology, (ii) allow targeted treatment and preventive measures and (iii) improve the prediction of long-term outcome. Classical attempts to define phenotypes have been one-dimensional, relying on few or single features such as triggers (exclusive viral wheeze vs. multiple trigger wheeze) or time course (early transient wheeze, persistent and late onset wheeze). These definitions are simple but essentially subjective. Recently, a multi-dimensional approach has been adopted. This approach is based on a wide range of features and relies on multivariate methods such as cluster or latent class analysis. Phenotypes identified in this manner are more complex but arguably more objective. Although phenotypes have an undisputed standing in current research on childhood asthma and wheezing, there is confusion about the meaning of the term 'phenotype' causing much circular debate. If phenotypes are meant to represent 'real' underlying disease entities rather than superficial features, there is a need for validation and harmonization of definitions. The multi-dimensional approach allows validation by replication across different populations and may contribute to a more reliable classification of childhood wheezing disorders and to improved precision of research relying on phenotype recognition, particularly in genetics. Ultimately, the underlying pathophysiology and aetiology will need to be understood to properly characterize the diseases causing recurrent wheeze in children.

Intravascular US-guided direct intrahepatic portocaval shunt with an expanded polytetrafluoroethylene-covered stent-graft

PURPOSE: To retrospectively evaluate the midterm patency rate of the nitinol (Viatorr, W.L. Gore and Associates, Flagstaff, Ariz) stent-graft for direct intrahepatic portacaval shunt (DIPS) creation. MATERIALS AND METHODS: Institutional Review Board approval for this retrospective HIPAA-compliant study was obtained with waiver of informed consent. DIPS was created in 18 men and one woman (median age, 54 years; range, 45-65 years) by using nitinol polytetrafluoroethylene (PTFE)-covered stent-grafts. The primary indications were intractable ascites (n = 14), acute variceal bleeding (n = 3), and hydrothorax (n = 2). Follow-up included Doppler ultrasonography at 1, 6, and 12 months and venography with manometry at 6-month intervals after the procedure. Shunt patency and cumulative survival were evaluated by using the Kaplan-Meier method and survival curves were plotted. Differences in mean portosystemic gradients (PSGs) were evaluated by using the Student t test. Multiple regression analysis for survival and DIPS patency were performed for the following parameters: Child-Pugh class, model of end-stage liver disease score, pre- and post-DIPS PSGs, pre-DIPS liver function tests, and pre-DIPS creatinine levels. RESULTS: DIPS creation was successful in all patients. Effective portal decompression and free antegrade shunt flow was achieved in all patients. Intraperitoneal bleeding occurred in one patient during the procedure and was controlled during the same procedure by placing a second nitinol stent-graft. The primary patency rate was 100% at all times during the follow-up period (range, 2 days to 30 months; mean, 256 days; median, 160 days). Flow restrictors were deployed in two (11%) of 19 patients. The 1-year mortality rate was 37% (seven of 19). CONCLUSION: Patency after DIPS creation with the nitinol PTFE-covered stent-graft was superior to that after TIPS with the nitinol stent-graft.

Estimating Diagnostic Test Accuracies for Brachyspira hyodysenteriae Accounting for the Complexities of Population Structure in Food Animals

For swine dysentery, which is caused by Brachyspira hyodysenteriae infection and is an economically important disease in intensive pig production systems worldwide, a perfect or error-free diagnostic test ("gold standard") is not available. In the absence of a gold standard, Bayesian latent class modelling is a well-established methodology for robust diagnostic test evaluation. In contrast to risk factor studies in food animals, where adjustment for within group correlations is both usual and required for good statistical practice, diagnostic test evaluation studies rarely take such clustering aspects into account, which can result in misleading results. The aim of the present study was to estimate test accuracies of a PCR originally designed for use as a confirmatory test, displaying a high diagnostic specificity, and cultural examination for B. hyodysenteriae. This estimation was conducted based on results of 239 samples from 103 herds originating from routine diagnostic sampling. Using Bayesian latent class modelling comprising of a hierarchical beta-binomial approach (which allowed prevalence across individual herds to vary as herd level random effect), robust estimates for the sensitivities of PCR and culture, as well as for the specificity of PCR, were obtained. The estimated diagnostic sensitivity of PCR (95% CI) and culture were 73.2% (62.3; 82.9) and 88.6% (74.9; 99.3), respectively. The estimated specificity of the PCR was 96.2% (90.9; 99.8). For test evaluation studies, a Bayesian latent class approach is well suited for addressing the considerable complexities of population structure in food animals.