Lifestyle Behaviors Associated With Body Fat Percent in 9- to 11-Year-Old Children.

PURPOSE: To examine (1) associations between body fat percent (BF) and lifestyle behaviors in children aged 9-11 years and (2) the consistency of these associations over a 10-year period. METHODS: In this repeat, cross-sectional study, 15,977 children aged 9-11 years completed an anthropometric assessment and the SportsLinx Lifestyle survey between 2004 and 2013. Body fat was estimated according to the sum of the triceps and subscapular skinfold measurements. Multilevel models were utilized to examine associations between BF and responses to the lifestyle survey while controlling for known covariates. RESULTS: Lifestyle behaviors explained 8.6% of the total variance in body fat. Specifically, negative associations were found between BF and active transport to school ( ß = -0.99 [0.19], P < .001), full-fat milk (-0.07 [0.15], P < .001), and sweetened beverage consumption (-0.40 [0.15], P = .007). Relative to the reference group of ≤8:00 PM, later bedtime was positively associated with BF: 8:00 to 8:59 PM ( ß = 1.60 [0.26], P < .001); 9:00 to 10:00 PM ( ß = 1.04 [0.24], P < .001); ≥10:00 PM ( ß = 1.18 [0.30], P < .001). Two-way interactions revealed opposing associations between BF and the consumption of low-calorie beverages for boys ( ß = 0.95 [0.25], P < .001) and girls ( ß = -0.85 [0.37], P = .021). There was no significant change in these associations over a 10-year period. CONCLUSIONS: In this population-level study covering a decade of data collection, lifestyle behaviors were associated with BF. Policies and interventions targeting population-level behavior change, such as active transport to school, sleep time, and consumption of full-fat milk, may offer an opportunity for improvements in BF.

Risk Factors for 1-Year Mortality and Hospital Utilization Patterns in Critical Care Survivors: A Retrospective, Observational, Population-Based Data Linkage Study.

OBJECTIVES: Clear understanding of the long-term consequences of critical care survivorship is essential. We investigated the care process and individual factors associated with long-term mortality among ICU survivors and explored hospital use in this group. DESIGN: Population-based data linkage study using the Secure Anonymised Information Linkage databank. SETTING: All ICUs between 2006 and 2013 in Wales, United Kingdom. PATIENTS: We identified 40,631 patients discharged alive from Welsh adult ICUs. INTERVENTIONS: None. MEASUREMENTS AND MAIN RESULTS: Primary outcome was 365-day survival. The secondary outcomes were 30- and 90-day survival and hospital utilization in the 365 days following ICU discharge. Kaplan-Meier curves were plotted to compare survival rates. Cox proportional hazards regression models were used to determine risk factors of mortality. Seven-thousand eight-hundred eighty-three patients (19.4%) died during the 1-year follow-up period. In the multivariable Cox regression analysis, advanced age and comorbidities were significant determinants of long-term mortality. Expedited discharge due to ICU bed shortage was associated with higher risk. The rate of hospitalization in the year prior to the critical care admission was 28 hospitalized days/1,000 d; post critical care was 88 hospitalized days/1,000 d for those who were still alive; and 57 hospitalized days/1,000 d and 412 hospitalized days/1,000 d for those who died by the end of the study, respectively. CONCLUSIONS: One in five ICU survivors die within 1 year, with advanced age and comorbidity being significant predictors of outcome, leading to high resource use. Care process factors indicating high system stress were associated with increased risk. More detailed understanding is needed on the effects of the potentially modifiable factors to optimize service delivery and improve long-term outcomes of the critically ill.

A Longitudinal Study of European Students' Alcohol Use and Related Behaviors as They Travel Abroad to Study.

BACKGROUND: Travelling away from home can be associated with fewer limits on behavior, particularly for students who participate in exchange programs. AIMS: To examine the effects of eight moderators on change in alcohol use and related negative outcomes, drug use and unprotected sexual behavior in European study abroad students before, during, and after their time abroad. METHODS: A three wave (before departure, while abroad, and after their return) longitudinal design collecting data on the frequency and volume of alcohol consumed, heavy episodic drinking, alcohol-related outcomes, drug use, and unprotected casual sex. RESULTS: The baseline survey was completed by 1145 students participating in one or two semester exchange programs (67.5% spent up to a semester abroad), of which 906 participated in two or more waves, representing 42 and 33 countries of origin and destination, respectively. Mean age was 22.2 years (SD = 2.28) and 72.7% were female. Students increased the amount of alcohol consumed by 35% (B = 0.32; 95% CI 0.287-0.349) and experienced more alcohol-related consequences (B = 0.15; 95% CI 0.089-0.219) during the study abroad experience, though levels fell below pre-departure levels when they returned home. Factors related to greater alcohol use while abroad include pre-departure expectations about alcohol use during the study abroad experience, psychological adjustment to the host country, academic involvement, and host country living costs. No statistically meaningful change in drug use and unprotected sexual behavior was observed. CONCLUSIONS: Studying abroad exposes European students to additional time-limited alcohol-related health risks.

The relationship between alcohol use and long-term cognitive decline in middle and late life: a longitudinal analysis using UK Biobank.

Background: Using UK Biobank data, this study sought to explain the causal relationship between alcohol intake and cognitive decline in middle and older aged populations. Methods: Data from 13 342 men and women, aged between 40 and 73 years were used in regression analysis that tested the functional relationship and impact of alcohol on cognitive performance. Performance was measured using mean reaction time (RT) and intra-individual variation (IIV) in RT, collected in response to a perceptual matching task. Covariates included body mass index, physical activity, tobacco use, socioeconomic status, education and baseline cognitive function. Results: A restricted cubic spline regression with three knots showed how the linear (ß1 = -0.048, 95% CI: -0.105 to -0.030) and non-linear effects (ß2 = 0.035, 95% CI: 0.007-0.059) of alcohol use on mean RT and IIV in RT (ß1 = -0.055, 95% CI: -0.125 to -0.034; ß2 = 0.034, 95% CI: 0.002-0.064) were significant adjusting for covariates. Cognitive function declined as alcohol use increased beyond 10 g/day. Decline was more apparent as age increased. Conclusions: The relationship between alcohol use and cognitive function is non-linear. Consuming more than one UK standard unit of alcohol per day is detrimental to cognitive performance and is more pronounced in older populations.

Object specificity and personal relevance in long-term visual remembering.

The personal relevance of an object is multi-faceted, each facet being capable of contributing to the effects on object memory attributed to personal relevance. An object's status as an individual object (object specificity), rather than just a category of object, is one such facet and its impact on the long-term visual remembering of everyday objects is assessed in two experiments. Images and drawings were produced under generic (e.g., "Please draw a bed") and personal exemplar (e.g., "Please draw your bed") instructions, and participants indicated the degree to which the image on which their drawing was based was of a specific object or a generic object. Object specificity induced a sense of time and place for a remembered object, the most recent encounter with the object being most salient. Other aspects of personal relevance collectively facilitated the retrieval of an object's category-irrelevant features (thereby increasing the vividness of the object image), the other objects with which it was seen, and a more general episodic sense of place. Against a broader theoretical perspective, it is proposed that visual episodic memory and visual knowledge are primary sources of information for specific personally relevant objects and generic objects, respectively.

Robust joint modeling of longitudinal measurements and time to event data using normal/independent distributions: a Bayesian approach.

Joint modeling of longitudinal data and survival data has been used widely for analyzing AIDS clinical trials, where a biological marker such as CD4 count measurement can be an important predictor of survival. In most of these studies, a normal distribution is used for modeling longitudinal responses, which leads to vulnerable inference in the presence of outliers in longitudinal measurements. Powerful distributions for robust analysis are normal/independent distributions, which include univariate and multivariate versions of the Student's t, the slash and the contaminated normal distributions in addition to the normal. In this paper, a linear-mixed effects model with normal/independent distribution for both random effects and residuals and Cox's model for survival time are used. For estimation, a Bayesian approach using Markov Chain Monte Carlo is adopted. Some simulation studies are performed for illustration of the proposed method. Also, the method is illustrated on a real AIDS data set and the best model is selected using some criteria.

A population-based investigation of behavioural and emotional problems and maternal mental health: associations with autism spectrum disorder and intellectual disability.

BACKGROUND: While research indicates elevated behavioural and emotional problems in children with autism spectrum disorders (ASD) and decreased well-being in their parents, studies do not typically separate out the contribution of ASD from that of associated intellectual disabilities (ID). We investigated child behavioural and emotional problems, and maternal mental health, among cases with and without ASD and ID in a large population-representative sample. METHODS: Cross-sectional comparison of child behavioural and emotional problems and maternal mental health measures among 18,415 children (5 to 16 years old), of whom 47 had an ASD, 51 combined ASD with ID, 590 had only ID, and the remainder were the comparison group with no ASD or ID. RESULTS: The prevalence of likely clinical levels of behavioural and emotional problems was highest among children with ASD (with and without ID). After controlling for age, gender, adversity, and maternal mental health, the presence of ASD and ID significantly and independently increased the odds for hyperactivity symptoms, conduct, and emotional problems. Emotional disorder was more prevalent in mothers of children with ASD (with or without ID). The presence of ASD, but not ID, significantly increased the odds for maternal emotional disorder. As has been found in previous research, positive maternal mental health was not affected by the presence of ASD or ID. CONCLUSIONS: ASD and ID are independent risk factors for behavioural and emotional problems. ASD (but not ID) is positively associated with maternal emotional disorder. Approaches to diagnosing hyperactivity and conduct problems in children with ASD may need to be reconsidered.

Object naming induces reliance on orientation-independent representations during longer-term, but not short-term, visual remembering.

Naming novel objects with novel count nouns changes how the objects are drawn from memory, revealing that object categorisation induces reliance on orientation-independent visual representations during longer-term remembering, but not during short-term remembering. Serial position effects integrate this finding with a more established conceptualisation of short-term and longer-term visual remembering in which the former is identified as keeping an item in mind. Adults were shown a series of four novel objects in orientations in which they would not normally be drawn from memory. When not named ("Look at this object"), the objects were drawn in the orientations in which they had been seen. When named with a novel count noun (e.g., "Look at this dax"), the final object continued to be depicted in the orientation in which it had been seen, but all other objects were depicted in an unseen but preferred (canonical) orientation, even though participants could still remember the orientations in which they had been seen. Although orientation-dependent exemplar representations appear to be more accessible than orientation-independent generic representations during short-term remembering, the reverse is the case during longer-term remembering. How the theoretical framework emerging from these observations accommodates a broader body of evidence is discussed.

Poverty transitions among families supporting a child with intellectual disability.

BACKGROUND: Little is known about child disability and dynamic aspects of poverty. METHOD: Analysis of data collected over a 12-month period for families (a) supporting a child with intellectual disability (n = 370), (b) supporting a child with other disability (n = 1,418), and (c) supporting a child with no disability (n = 7,215). RESULTS: When compared to families not supporting a child with disability, families supporting a child with intellectual disability were (a) more likely to be poor, (b) more likely to become poor, (c) less likely to escape from being poor. Half of poverty transitions were associated with identifiable potential trigger events. CONCLUSIONS: There were few differences between families supporting or not supporting a child with disability with regard to either levels of exposure to potential trigger events or to the strength of the association between exposure and poverty transitions.

A longitudinal analysis of teenage drinking patterns.

This paper examines the multinomial recurrent outcome 'teenage drinking' within a statistical modeling paradigm. The proposed statistical modeling relates drinking to a set of explanatory variables, which include subjective as well as objective measures. In order to assess the degree to which explanatory variables influence smoking, an adequate statistical model must handle the possibility that substantial variation between respondents will be due to omitted variables, multicollinearity and past behavior. To address these issues, longitudinal data using appropriate statistical modeling are essential. In this paper we demonstrate the application of appropriate longitudinal modeling using data from the Yorkshire (UK) Health Related Behavior Questionnaire. The results suggest there may be substantial heterogeneity due to omitted variables in the data and complex inter-relationships between observed explanatory variables. For example, social cultures of drinking and parental involvement appear to influence heavy drinking and drug use.

Longitudinal access and exposure to green-blue spaces and individual-level mental health and well-being: protocol for a longitudinal, population-wide record-linked natural experiment.

INTRODUCTION: Studies suggest that access and exposure to green-blue spaces (GBS) have beneficial impacts on mental health. However, the evidence base is limited with respect to longitudinal studies. The main aim of this longitudinal, population-wide, record-linked natural experiment, is to model the daily lived experience by linking GBS accessibility indices, residential GBS exposure and health data; to enable quantification of the impact of GBS on well-being and common mental health disorders, for a national population. METHODS AND ANALYSIS: This research will estimate the impact of neighbourhood GBS access, GBS exposure and visits to GBS on the risk of common mental health conditions and the opportunity for promoting subjective well-being (SWB); both key priorities for public health. We will use a Geographic Information System (GIS) to create quarterly household GBS accessibility indices and GBS exposure using digital map and satellite data for 1.4 million homes in Wales, UK (2008-2018). We will link the GBS accessibility indices and GBS exposures to individual-level mental health outcomes for 1.7 million people with general practitioner (GP) data and data from the National Survey for Wales (n=~12 000) on well-being in the Secure Anonymised Information Linkage (SAIL) Databank. We will examine if these associations are modified by multiple sociophysical variables, migration and socioeconomic disadvantage. Subgroup analyses will examine associations by different types of GBS. This longitudinal study will be augmented by cross-sectional research using survey data on self-reported visits to GBS and SWB. ETHICS AND DISSEMINATION: All data will be anonymised and linked within the privacy protecting SAIL Databank. We will be using anonymised data and therefore we are exempt from National Research Ethics Committee (NREC). An Information Governance Review Panel (IGRP) application (Project ID: 0562) to link these data has been approved.The research programme will be undertaken in close collaboration with public/patient involvement groups. A multistrategy programme of dissemination is planned with the academic community, policy-makers, practitioners and the public.

Do doctors' attachment styles and emotional intelligence influence patients' emotional expressions in primary care consultations? An exploratory study using multilevel analysis.

OBJECTIVE: To investigate whether and how doctors' attachment styles and emotional intelligence (EI) might influence patients' emotional expressions in general practice consultations. METHODS: Video recordings of 26 junior doctors consulting with 173 patients were coded using the Verona Coding Definition of Emotional Sequences (VR-CoDES). Doctors' attachment style was scored across two dimensions, avoidance and anxiety, using the Experiences in Close Relationships: Short Form questionnaire. EI was assessed with the Mayer-Salovey-Caruso Emotional Intelligence Test. Multilevel Poisson regressions modelled the probability of patients' expressing emotional distress, considering doctors' attachment styles and EI and demographic and contextual factors. RESULTS: Both attachment styles and EI were significantly associated with frequency of patients' cues, with patient- and doctor-level explanatory variables accounting for 42% of the variance in patients' cues. The relative contribution of attachment styles and EI varied depending on whether patients' presenting complaints were physical or psychosocial in nature. CONCLUSION: Doctors' attachment styles and levels of EI are associated with patients' emotional expressions in primary care consultations. Further research is needed to investigate how these two variables interact and influence provider responses and patient outcomes. PRACTICE IMPLICATIONS: Understanding how doctors' psychological characteristics influence PPC may help to optimise undergraduate and postgraduate medical education.

Emergency hospital admissions associated with a non-randomised housing intervention meeting national housing quality standards: a longitudinal data linkage study.

BACKGROUND: We investigated tenant healthcare utilisation associated with upgrading 8558 council houses to a national quality standard. Homes received multiple internal and external improvements and were analysed using repeated measures of healthcare utilisation. METHODS: The primary outcome was emergency hospital admissions for cardiorespiratory conditions and injuries for residents aged 60 years and over. Secondary outcomes included each of the separate conditions, for tenants aged 60 and over, and for all ages. Council home address and intervention records for eight housing cointerventions were anonymously linked to demographic data, hospital admissions and deaths for individuals in a dynamic cohort. Counts of health events were analysed using multilevel regression models to investigate associations between receipt of each housing improvement, adjusting for potential confounding factors and regional trends. RESULTS: Residents aged 60 years and over living in homes when improvements were made were associated with up to 39% fewer admissions compared with those living in homes that were not upgraded (incidence rate ratio=0.61, 95% CI 0.53 to 0.72). Reduced admissions were associated with electrical systems, windows and doors, wall insulation, and garden paths. There were small non-significant reductions for the primary outcome associated with upgrading heating, adequate loft insulation, new kitchens and new bathrooms. CONCLUSION: Results suggest that hospital admissions can be avoided through improving whole home quality standards. This is the first large-scale longitudinal evaluation of a whole home intervention that has evaluated multiple improvement elements using individual-level objective routine health data.

Study protocol for investigating the impact of community home modification services on hospital utilisation for fall injuries: a controlled longitudinal study using data linkage.

INTRODUCTION: This study will evaluate the effectiveness of home adaptations, both in preventing hospital admissions due to falls for older people, and improving timely discharge. Results will provide evidence for services at the interface between health and social care, informing policies seeking to promote healthy ageing through prudent healthcare and fall prevention. METHODS AND ANALYSIS: All individuals living in Wales, UK, aged 60 years and over, will be included in the study using anonymised linked data from the Secure Anonymised Information Linkage Databank. We will use a national database of home modifications implemented by the charity organisation Care & Repair Cymru (C&R) from 2009 to 2017 to define an intervention cohort. We will use the electronic Frailty Index to assign individual levels of frailty (fit, mild, moderate or severe) and use these to create a comparator group (non-C&R) of people who have not received a C&R intervention. Coprimary outcomes will be quarterly numbers of emergency hospital admissions attributed to falls at home, and the associated length of stay. Secondary outcomes include the time in moving to a care home following a fall, and the indicative financial costs of care for individuals who had a fall. We will use appropriate multilevel generalised linear models to analyse the number of hospital admissions related to falls. We will use Cox proportional hazard models to compare the length of stay for fall-related hospital admissions and the time in moving to a care home between the C&R and non-C&R cohorts. We will assess the impact per frailty group, correct for population migration and adjust for confounding variables. Indicative costs will be calculated using financial codes for individual-level hospital stays. Results will provide evidence for services at the interface between health and social care, informing policies seeking to promote healthy ageing through prudent healthcare and prevention. ETHICS AND DISSEMINATION: Information governance requirements for the use of record-linked data have been approved and only anonymised data will be used in our analysis. Our results will be submitted for publication in peer-reviewed journals. We will also work with lay members and the knowledge transfer team at Swansea University to create communication and dissemination materials on key findings.

Using Neural Networks with Routine Health Records to Identify Suicide Risk: Feasibility Study.

BACKGROUND: Each year, approximately 800,000 people die by suicide worldwide, accounting for 1-2 in every 100 deaths. It is always a tragic event with a huge impact on family, friends, the community and health professionals. Unfortunately, suicide prevention and the development of risk assessment tools have been hindered by the complexity of the underlying mechanisms and the dynamic nature of a person's motivation and intent. Many of those who die by suicide had contact with health services in the preceding year but identifying those most at risk remains a challenge. OBJECTIVE: To explore the feasibility of using artificial neural networks with routinely collected electronic health records to support the identification of those at high risk of suicide when in contact with health services. METHODS: Using the Secure Anonymised Information Linkage Databank UK, we extracted the data of those who died by suicide between 2001 and 2015 and paired controls. Looking at primary (general practice) and secondary (hospital admissions) electronic health records, we built a binary feature vector coding the presence of risk factors at different times prior to death. Risk factors included: general practice contact and hospital admission; diagnosis of mental health issues; injury and poisoning; substance misuse; maltreatment; sleep disorders; and the prescription of opiates and psychotropics. Basic artificial neural networks were trained to differentiate between the suicide cases and paired controls. We interpreted the output score as the estimated suicide risk. System performance was assessed with 10x10-fold repeated cross-validation, and its behavior was studied by representing the distribution of estimated risk across the cases and controls, and the distribution of factors across estimated risks. RESULTS: We extracted a total of 2604 suicide cases and 20 paired controls per case. Our best system attained a mean error rate of 26.78% (SD 1.46; 64.57% of sensitivity and 81.86% of specificity). While the distribution of controls was concentrated around estimated risks < 0.5, cases were almost uniformly distributed between 0 and 1. Prescription of psychotropics, depression and anxiety, and self-harm increased the estimated risk by ~0.4. At least 95% of those presenting these factors were identified as suicide cases. CONCLUSIONS: Despite the simplicity of the implemented system, the proposed methodology obtained an accuracy like other published methods based on specialized questionnaire generated data. Most of the errors came from the heterogeneity of patterns shown by suicide cases, some of which were identical to those of the paired controls. Prescription of psychotropics, depression and anxiety, and self-harm were strongly linked with higher estimated risk scores, followed by hospital admission and long-term drug and alcohol misuse. Other risk factors like sleep disorders and maltreatment had more complex effects.

Premature mortality among people with severe mental illness - New evidence from linked primary care data.

Studies assessing premature mortality in people with severe mental illness (SMI) are usually based in one setting, hospital (secondary care inpatients and/or outpatients) or community (primary care). This may lead to ascertainment bias. This study aimed to estimate standardised mortality ratios (SMRs) for all-cause and cause-specific mortality in people with SMI drawn from linked primary and secondary care populations compared to the general population. SMRs were calculated using the indirect method for a United Kingdom population of almost four million between 2004 and 2013. The all-cause SMR was higher in the cohort identified from secondary care hospital admissions (SMR: 2.9; 95% CI: 2.8-3.0) than from primary care (SMR: 2.2; 95% CI: 2.1-2.3) when compared to the general population. The SMR for the combined cohort was 2.6 (95% CI: 2.5-2.6). Cause specific SMRs in the combined cohort were particularly elevated in those with SMI relative to the general population for ill-defined and unknown causes, suicide, substance abuse, Parkinson's disease, accidents, dementia, infections and respiratory disorders (particularly pneumonia), and Alzheimer's disease. Solely hospital admission based studies, which have dominated the literature hitherto, somewhat over-estimate premature mortality in those with SMI. People with SMI are more likely to die by ill-defined and unknown causes, suicide and other less common and often under-reported causes. Comprehensive characterisation of mortality is important to inform policy and practice and to discriminate settings to allow for proportionate interventions to address this health injustice.

Creating individual level air pollution exposures in an anonymised data safe haven: a platform for evaluating impact on educational attainment.

INTRODUCTION: There is a lack of evidence on the adverse effects of air pollution on cognition for people with air quality-related health conditions. We propose that educational attainment, as a proxy for cognition, may increase with improved air quality. This study will explore whether asthma and seasonal allergic rhinitis, when exacerbated by acute exposure to air pollution, is associated with educational attainment. OBJECTIVE: To describe the preparation of individual and household-level linked environmental and health data for analysis within an anonymised safe haven. Also to introduce our statistical analysis plan for our study: COgnition, Respiratory Tract illness and Effects of eXposure (CORTEX). METHODS: We imported daily air pollution and aeroallergen data, and individual level education data into the SAIL databank, an anonymised safe haven for person-based records. We linked individual-level education, socioeconomic and health data to air quality data for home and school locations, creating tailored exposures for individuals across a city. We developed daily exposure data for all pupils in repeated cross sectional exam cohorts (2009-2015). CONCLUSION: We have used the SAIL databank, an innovative, data safe haven to create individual-level exposures to air pollution and pollen for multiple daily home and school locations. The analysis platform will allow us to evaluate retrospectively the impact of air quality on attainment for multiple cross-sectional cohorts of pupils. Our methods will allow us to distinguish between the pollution impacts on educational attainment for pupils with and without respiratory health conditions. The results from this study will further our understanding of the effects of air quality and respiratory-related health conditions on cognition. HIGHLIGHTS: This city-wide study includes longitudinal routinely-recorded educational attainment data for all pupils taking exams over seven years;High spatial resolution air pollution data were linked within a privacy protected databank to obtain individual exposure at multiple daily locations;This study will use health data linked at the individual level to explore associations between air pollution, related morbidity, and educational attainment.

Teenage smoking: a longitudinal analysis.

This paper examines the binary recurrent outcome "teenage smoking" within a statistical modelling paradigm. The proposed statistical modelling relates smoking to a set of explanatory variables, which include subjective as well as objective measures. In order to assess the degree to which explanatory variables influence smoking, an adequate statistical model must handle the possibility that substantial variation between respondents will be due to omitted variables, multicollinearity and past behaviour. An earlier paper, using a secondary cross-sectional data source, concluded that an investigation of smoking needs to be based on longitudinal data using appropriate statistical modelling. The same data source provided observations on young adults over a period of 2 years. For comparison purposes, the same cross-sectional model was fitted to the longitudinal data. The results suggest there may be substantial heterogeneity due to omitted variables in the data and complex inter-relationships between observed explanatory variables leading to underestimation. Longitudinal data provide additional flexibility to control for omitted variables and are necessary to investigate dynamic social processes such as smoking. The results from our analysis suggest that the effects of variables reported in the literature on teenage smoking may be overestimated. For example, the role of peer pressure may not be as clear as it has been made out to be.

Robust modeling of differential gene expression data using normal/independent distributions: a Bayesian approach.

In this paper, the problem of identifying differentially expressed genes under different conditions using gene expression microarray data, in the presence of outliers, is discussed. For this purpose, the robust modeling of gene expression data using some powerful distributions known as normal/independent distributions is considered. These distributions include the Student's t and normal distributions which have been used previously, but also include extensions such as the slash, the contaminated normal and the Laplace distributions. The purpose of this paper is to identify differentially expressed genes by considering these distributional assumptions instead of the normal distribution. A Bayesian approach using the Markov Chain Monte Carlo method is adopted for parameter estimation. Two publicly available gene expression data sets are analyzed using the proposed approach. The use of the robust models for detecting differentially expressed genes is investigated. This investigation shows that the choice of model for differentiating gene expression data is very important. This is due to the small number of replicates for each gene and the existence of outlying data. Comparison of the performance of these models is made using different statistical criteria and the ROC curve. The method is illustrated using some simulation studies. We demonstrate the flexibility of these robust models in identifying differentially expressed genes.