Addressing issues in sparseness, ecological bias and formulation of the adjacency matrix in Bayesian spatio-temporal analysis of disease counts

The main objective of this PhD was to further develop Bayesian spatio-temporal models (specifically the Conditional Autoregressive (CAR) class of models), for the analysis of sparse disease outcomes such as birth defects. The motivation for the thesis arose from problems encountered when analyzing a large birth defect registry in New South Wales. The specific components and related research objectives of the thesis were developed from gaps in the literature on current formulations of the CAR model, and health service planning requirements. Data from a large probabilistically-linked database from 1990 to 2004, consisting of fields from two separate registries: the Birth Defect Registry (BDR) and Midwives Data Collection (MDC) were used in the analyses in this thesis. The main objective was split into smaller goals. The first goal was to determine how the specification of the neighbourhood weight matrix will affect the smoothing properties of the CAR model, and this is the focus of chapter 6. Secondly, I hoped to evaluate the usefulness of incorporating a zero-inflated Poisson (ZIP) component as well as a shared-component model in terms of modeling a sparse outcome, and this is carried out in chapter 7. The third goal was to identify optimal sampling and sample size schemes designed to select individual level data for a hybrid ecological spatial model, and this is done in chapter 8. Finally, I wanted to put together the earlier improvements to the CAR model, and along with demographic projections, provide forecasts for birth defects at the SLA level. Chapter 9 describes how this is done. For the first objective, I examined a series of neighbourhood weight matrices, and showed how smoothing the relative risk estimates according to similarity by an important covariate (i.e. maternal age) helped improve the model’s ability to recover the underlying risk, as compared to the traditional adjacency (specifically the Queen) method of applying weights. Next, to address the sparseness and excess zeros commonly encountered in the analysis of rare outcomes such as birth defects, I compared a few models, including an extension of the usual Poisson model to encompass excess zeros in the data. This was achieved via a mixture model, which also encompassed the shared component model to improve on the estimation of sparse counts through borrowing strength across a shared component (e.g. latent risk factor/s) with the referent outcome (caesarean section was used in this example). Using the Deviance Information Criteria (DIC), I showed how the proposed model performed better than the usual models, but only when both outcomes shared a strong spatial correlation. The next objective involved identifying the optimal sampling and sample size strategy for incorporating individual-level data with areal covariates in a hybrid study design. I performed extensive simulation studies, evaluating thirteen different sampling schemes along with variations in sample size. This was done in the context of an ecological regression model that incorporated spatial correlation in the outcomes, as well as accommodating both individual and areal measures of covariates. Using the Average Mean Squared Error (AMSE), I showed how a simple random sample of 20% of the SLAs, followed by selecting all cases in the SLAs chosen, along with an equal number of controls, provided the lowest AMSE. The final objective involved combining the improved spatio-temporal CAR model with population (i.e. women) forecasts, to provide 30-year annual estimates of birth defects at the Statistical Local Area (SLA) level in New South Wales, Australia. The projections were illustrated using sixteen different SLAs, representing the various areal measures of socio-economic status and remoteness. A sensitivity analysis of the assumptions used in the projection was also undertaken. By the end of the thesis, I will show how challenges in the spatial analysis of rare diseases such as birth defects can be addressed, by specifically formulating the neighbourhood weight matrix to smooth according to a key covariate (i.e. maternal age), incorporating a ZIP component to model excess zeros in outcomes and borrowing strength from a referent outcome (i.e. caesarean counts). An efficient strategy to sample individual-level data and sample size considerations for rare disease will also be presented. Finally, projections in birth defect categories at the SLA level will be made.

Brain-derived neurotrophic factor (BDNF) gene : no major impact on antidepressant treatment response

The brain-derived neurotrophic factor (BDNF) has been suggested to play a pivotal role in the aetiology of affective disorders. In order to further clarify the impact of BDNF gene variation on major depression as well as antidepressant treatment response, association of three BDNF polymorphisms [rs7103411, Val66Met (rs6265) and rs7124442] with major depression and antidepressant treatment response was investigated in an overall sample of 268 German patients with major depression and 424 healthy controls. False discovery rate (FDR) was applied to control for multiple testing. Additionally, ten markers in BDNF were tested for association with citalopram outcome in the STAR*D sample. While BDNF was not associated with major depression as a categorical diagnosis, the BDNF rs7124442 TT genotype was significantly related to worse treatment outcome over 6 wk in major depression (p=0.01) particularly in anxious depression (p=0.003) in the German sample. However, BDNF rs7103411 and rs6265 similarly predicted worse treatment response over 6 wk in clinical subtypes of depression such as melancholic depression only (rs7103411: TTcovariates. The STAR*D analyses did not yield significant results at any of the ten BDNF markers. Our results do not support an association between genetic variation in BDNF and antidepressant treatment response or remission. Post-hoc analyses provide some preliminary support for a potential minor role of genetic variation in BDNF and antidepressant treatment outcome in the context of melancholic depression.

Bayesian models for longitudinal data

Longitudinal data, where data are repeatedly observed or measured on a temporal basis of time or age provides the foundation of the analysis of processes which evolve over time, and these can be referred to as growth or trajectory models. One of the traditional ways of looking at growth models is to employ either linear or polynomial functional forms to model trajectory shape, and account for variation around an overall mean trend with the inclusion of random eects or individual variation on the functional shape parameters. The identification of distinct subgroups or sub-classes (latent classes) within these trajectory models which are not based on some pre-existing individual classification provides an important methodology with substantive implications. The identification of subgroups or classes has a wide application in the medical arena where responder/non-responder identification based on distinctly diering trajectories delivers further information for clinical processes. This thesis develops Bayesian statistical models and techniques for the identification of subgroups in the analysis of longitudinal data where the number of time intervals is limited. These models are then applied to a single case study which investigates the neuropsychological cognition for early stage breast cancer patients undergoing adjuvant chemotherapy treatment from the Cognition in Breast Cancer Study undertaken by the Wesley Research Institute of Brisbane, Queensland. Alternative formulations to the linear or polynomial approach are taken which use piecewise linear models with a single turning point, change-point or knot at a known time point and latent basis models for the non-linear trajectories found for the verbal memory domain of cognitive function before and after chemotherapy treatment. Hierarchical Bayesian random eects models are used as a starting point for the latent class modelling process and are extended with the incorporation of covariates in the trajectory profiles and as predictors of class membership. The Bayesian latent basis models enable the degree of recovery post-chemotherapy to be estimated for short and long-term followup occasions, and the distinct class trajectories assist in the identification of breast cancer patients who maybe at risk of long-term verbal memory impairment.

Hierarchical models for 2D presence/absence data having ambiguous zeroes: With a biogeographical case study on dingo behaviour

This dissertation is primarily an applied statistical modelling investigation, motivated by a case study comprising real data and real questions. Theoretical questions on modelling and computation of normalization constants arose from pursuit of these data analytic questions. The essence of the thesis can be described as follows. Consider binary data observed on a two-dimensional lattice. A common problem with such data is the ambiguity of zeroes recorded. These may represent zero response given some threshold (presence) or that the threshold has not been triggered (absence). Suppose that the researcher wishes to estimate the effects of covariates on the binary responses, whilst taking into account underlying spatial variation, which is itself of some interest. This situation arises in many contexts and the dingo, cypress and toad case studies described in the motivation chapter are examples of this. Two main approaches to modelling and inference are investigated in this thesis. The first is frequentist and based on generalized linear models, with spatial variation modelled by using a block structure or by smoothing the residuals spatially. The EM algorithm can be used to obtain point estimates, coupled with bootstrapping or asymptotic MLE estimates for standard errors. The second approach is Bayesian and based on a three- or four-tier hierarchical model, comprising a logistic regression with covariates for the data layer, a binary Markov Random field (MRF) for the underlying spatial process, and suitable priors for parameters in these main models. The three-parameter autologistic model is a particular MRF of interest. Markov chain Monte Carlo (MCMC) methods comprising hybrid Metropolis/Gibbs samplers is suitable for computation in this situation. Model performance can be gauged by MCMC diagnostics. Model choice can be assessed by incorporating another tier in the modelling hierarchy. This requires evaluation of a normalization constant, a notoriously difficult problem. Difficulty with estimating the normalization constant for the MRF can be overcome by using a path integral approach, although this is a highly computationally intensive method. Different methods of estimating ratios of normalization constants (N Cs) are investigated, including importance sampling Monte Carlo (ISMC), dependent Monte Carlo based on MCMC simulations (MCMC), and reverse logistic regression (RLR). I develop an idea present though not fully developed in the literature, and propose the Integrated mean canonical statistic (IMCS) method for estimating log NC ratios for binary MRFs. The IMCS method falls within the framework of the newly identified path sampling methods of Gelman & Meng (1998) and outperforms ISMC, MCMC and RLR. It also does not rely on simplifying assumptions, such as ignoring spatio-temporal dependence in the process. A thorough investigation is made of the application of IMCS to the three-parameter Autologistic model. This work introduces background computations required for the full implementation of the four-tier model in Chapter 7. Two different extensions of the three-tier model to a four-tier version are investigated. The first extension incorporates temporal dependence in the underlying spatio-temporal process. The second extensions allows the successes and failures in the data layer to depend on time. The MCMC computational method is extended to incorporate the extra layer. A major contribution of the thesis is the development of a fully Bayesian approach to inference for these hierarchical models for the first time. Note: The author of this thesis has agreed to make it open access but invites people downloading the thesis to send her an email via the 'Contact Author' function.

Development of accident modification factors on intersection crash maneuvers

In order to estimate the safety impact of roadway interventions engineers need to collect, analyze, and interpret the results of carefully implemented data collection efforts. The intent of these studies is to develop Accident Modification Factors (AMF's), which are used to predict the safety impact of various road safety features at other locations or in upon future enhancements. Models are typically estimated to estimate AMF's for total crashes, but can and should be estimated for crash outcomes as well. This paper first describes data collected with the intent estimate AMF's for rural intersections in the state of Georgia within the United Sates. Modeling results of crash prediction models for the crash outcomes: angle, head-on, rear-end, sideswipe (same direction and opposite direction) and pedestrian-involved crashes are then presented and discussed. The analysis reveals that factors such as the Annual Average Daily Traffic (AADT), the presence of turning lanes, and the number of driveways have a positive association with each type of crash, while the median width and the presence of lighting are negatively associated with crashes. The model covariates are related to crash outcome in different ways, suggesting that crash outcomes are associated with different pre-crash conditions.

On the nature of over-dispersion in motor vehicle crash prediction models

Statistical modeling of traffic crashes has been of interest to researchers for decades. Over the most recent decade many crash models have accounted for extra-variation in crash counts—variation over and above that accounted for by the Poisson density. The extra-variation – or dispersion – is theorized to capture unaccounted for variation in crashes across sites. The majority of studies have assumed fixed dispersion parameters in over-dispersed crash models—tantamount to assuming that unaccounted for variation is proportional to the expected crash count. Miaou and Lord [Miaou, S.P., Lord, D., 2003. Modeling traffic crash-flow relationships for intersections: dispersion parameter, functional form, and Bayes versus empirical Bayes methods. Transport. Res. Rec. 1840, 31–40] challenged the fixed dispersion parameter assumption, and examined various dispersion parameter relationships when modeling urban signalized intersection accidents in Toronto. They suggested that further work is needed to determine the appropriateness of the findings for rural as well as other intersection types, to corroborate their findings, and to explore alternative dispersion functions. This study builds upon the work of Miaou and Lord, with exploration of additional dispersion functions, the use of an independent data set, and presents an opportunity to corroborate their findings. Data from Georgia are used in this study. A Bayesian modeling approach with non-informative priors is adopted, using sampling-based estimation via Markov Chain Monte Carlo (MCMC) and the Gibbs sampler. A total of eight model specifications were developed; four of them employed traffic flows as explanatory factors in mean structure while the remainder of them included geometric factors in addition to major and minor road traffic flows. The models were compared and contrasted using the significance of coefficients, standard deviance, chi-square goodness-of-fit, and deviance information criteria (DIC) statistics. The findings indicate that the modeling of the dispersion parameter, which essentially explains the extra-variance structure, depends greatly on how the mean structure is modeled. In the presence of a well-defined mean function, the extra-variance structure generally becomes insignificant, i.e. the variance structure is a simple function of the mean. It appears that extra-variation is a function of covariates when the mean structure (expected crash count) is poorly specified and suffers from omitted variables. In contrast, when sufficient explanatory variables are used to model the mean (expected crash count), extra-Poisson variation is not significantly related to these variables. If these results are generalizable, they suggest that model specification may be improved by testing extra-variation functions for significance. They also suggest that known influences of expected crash counts are likely to be different than factors that might help to explain unaccounted for variation in crashes across sites

Effects of transportation accessibility on residential property values: Application of spatial hedonic price model in Seoul, South Korea, metropolitan area

A number of studies have focused on estimating the effects of accessibility on housing values by using the hedonic price model. In the majority of studies, estimation results have revealed that housing values increase as accessibility improves, although the magnitude of estimates has varied across studies. Adequately estimating the relationship between transportation accessibility and housing values is challenging for at least two reasons. First, the monocentric city assumption applied in location theory is no longer valid for many large or growing cities. Second, rather than being randomly distributed in space, housing values are clustered in space—often exhibiting spatial dependence. Recognizing these challenges, a study was undertaken to develop a spatial lag hedonic price model in the Seoul, South Korea, metropolitan region, which includes a measure of local accessibility as well as systemwide accessibility, in addition to other model covariates. Although the accessibility measures can be improved, the modeling results suggest that the spatial interactions of apartment sales prices occur across and within traffic analysis zones, and the sales prices for apartment communities are devalued as accessibility deteriorates. Consistent with findings in other cities, this study revealed that the distance to the central business district is still a significant determinant of sales price.

Remaining useful life prediction using elliptical basis function network and Markov Chain

This paper presents a novel method for remaining useful life prediction using the Elliptical Basis Function (EBF) network and a Markov chain. The EBF structure is trained by a modified Expectation-Maximization (EM) algorithm in order to take into account the missing covariate set. No explicit extrapolation is needed for internal covariates while a Markov chain is constructed to represent the evolution of external covariates in the study. The estimated external and the unknown internal covariates constitute an incomplete covariate set which are then used and analyzed by the EBF network to provide survival information of the asset. It is shown in the case study that the method slightly underestimates the remaining useful life of an asset which is a desirable result for early maintenance decision and resource planning.

Probabilistic subgroup identification using Bayesian finite mixture modelling : a case study in Parkinson’s disease phenotype identification

This article explores the use of probabilistic classification, namely finite mixture modelling, for identification of complex disease phenotypes, given cross-sectional data. In particular, if focuses on posterior probabilities of subgroup membership, a standard output of finite mixture modelling, and how the quantification of uncertainty in these probabilities can lead to more detailed analyses. Using a Bayesian approach, we describe two practical uses of this uncertainty: (i) as a means of describing a person’s membership to a single or multiple latent subgroups and (ii) as a means of describing identified subgroups by patient-centred covariates not included in model estimation. These proposed uses are demonstrated on a case study in Parkinson’s disease (PD), where latent subgroups are identified using multiple symptoms from the Unified Parkinson’s Disease Rating Scale (UPDRS).

Malnutrition and clinical outcomes in elderly patients from a Singapore acute hospital

Older adults, especially those acutely ill, are vulnerable to developing malnutrition due to a range of risk factors. The high prevalence and extensive consequences of malnutrition in hospitalised older adults have been reported extensively. However, there are few well-designed longitudinal studies that report the independent relationship between malnutrition and clinical outcomes after adjustment for a wide range of covariates. Acutely ill older adults are exceptionally prone to nutritional decline during hospitalisation, but few reports have studied this change and impact on clinical outcomes. In the rapidly ageing Singapore population, all this evidence is lacking, and the characteristics associated with the risk of malnutrition are also not well-documented. Despite the evidence on malnutrition prevalence, it is often under-recognised and under-treated. It is therefore crucial that validated nutrition screening and assessment tools are used for early identification of malnutrition. Although many nutrition screening and assessment tools are available, there is no universally accepted method for defining malnutrition risk and nutritional status. Most existing tools have been validated amongst Caucasians using various approaches, but they are rarely reported in the Asian elderly and none has been validated in Singapore. Due to the multiethnicity, cultural, and language differences in Singapore older adults, the results from non-Asian validation studies may not be applicable. Therefore it is important to identify validated population and setting specific nutrition screening and assessment methods to accurately detect and diagnose malnutrition in Singapore. The aims of this study are therefore to: i) characterise hospitalised elderly in a Singapore acute hospital; ii) describe the extent and impact of admission malnutrition; iii) identify and evaluate suitable methods for nutritional screening and assessment; and iv) examine changes in nutritional status during admission and their impact on clinical outcomes. A total of 281 participants, with a mean (+SD) age of 81.3 (+7.6) years, were recruited from three geriatric wards in Tan Tock Seng Hospital over a period of eight months. They were predominantly Chinese (83%) and community-dwellers (97%). They were screened within 72 hours of admission by a single dietetic technician using four nutrition screening tools [Tan Tock Seng Hospital Nutrition Screening Tool (TTSH NST), Nutritional Risk Screening 2002 (NRS 2002), Mini Nutritional Assessment-Short Form (MNA-SF), and Short Nutritional Assessment Questionnaire (SNAQ©)] that were administered in no particular order. The total scores were not computed during the screening process so that the dietetic technician was blinded to the results of all the tools. Nutritional status was assessed by a single dietitian, who was blinded to the screening results, using four malnutrition assessment methods [Subjective Global Assessment (SGA), Mini Nutritional Assessment (MNA), body mass index (BMI), and corrected arm muscle area (CAMA)]. The SGA rating was completed prior to computation of the total MNA score to minimise bias. Participants were reassessed for weight, arm anthropometry (mid-arm circumference, triceps skinfold thickness), and SGA rating at discharge from the ward. The nutritional assessment tools and indices were validated against clinical outcomes (length of stay (LOS) >11days, discharge to higher level care, 3-month readmission, 6-month mortality, and 6-month Modified Barthel Index) using multivariate logistic regression. The covariates included age, gender, race, dementia (defined using DSM IV criteria), depression (defined using a single question “Do you often feel sad or depressed?”), severity of illness (defined using a modified version of the Severity of Illness Index), comorbidities (defined using Charlson Comorbidity Index, number of prescribed drugs and admission functional status (measured using Modified Barthel Index; MBI). The nutrition screening tools were validated against the SGA, which was found to be the most appropriate nutritional assessment tool from this study (refer section 5.6) Prevalence of malnutrition on admission was 35% (defined by SGA), and it was significantly associated with characteristics such as swallowing impairment (malnourished vs well-nourished: 20% vs 5%), poor appetite (77% vs 24%), dementia (44% vs 28%), depression (34% vs 22%), and poor functional status (MBI 48.3+29.8 vs 65.1+25.4). The SGA had the highest completion rate (100%) and was predictive of the highest number of clinical outcomes: LOS >11days (OR 2.11, 95% CI [1.17- 3.83]), 3-month readmission (OR 1.90, 95% CI [1.05-3.42]) and 6-month mortality (OR 3.04, 95% CI [1.28-7.18]), independent of a comprehensive range of covariates including functional status, disease severity and cognitive function. SGA is therefore the most appropriate nutritional assessment tool for defining malnutrition. The TTSH NST was identified as the most suitable nutritional screening tool with the best diagnostic performance against the SGA (AUC 0.865, sensitivity 84%, specificity 79%). Overall, 44% of participants experienced weight loss during hospitalisation, and 27% had weight loss >1% per week over median LOS 9 days (range 2-50). Wellnourished (45%) and malnourished (43%) participants were equally prone to experiencing decline in nutritional status (defined by weight loss >1% per week). Those with reduced nutritional status were more likely to be discharged to higher level care (adjusted OR 2.46, 95% CI [1.27-4.70]). This study is the first to characterise malnourished hospitalised older adults in Singapore. It is also one of the very few studies to (a) evaluate the association of admission malnutrition with clinical outcomes in a multivariate model; (b) determine the change in their nutritional status during admission; and (c) evaluate the validity of nutritional screening and assessment tools amongst hospitalised older adults in an Asian population. Results clearly highlight that admission malnutrition and deterioration in nutritional status are prevalent and are associated with adverse clinical outcomes in hospitalised older adults. With older adults being vulnerable to risks and consequences of malnutrition, it is important that they are systematically screened so timely and appropriate intervention can be provided. The findings highlighted in this thesis provide an evidence base for, and confirm the validity of the current nutrition screening and assessment tools used among hospitalised older adults in Singapore. As the older adults may have developed malnutrition prior to hospital admission, or experienced clinically significant weight loss of >1% per week of hospitalisation, screening of the elderly should be initiated in the community and continuous nutritional monitoring should extend beyond hospitalisation.

Prospective association between physical activity and falls in community-dwelling older women

Abstract Objective: To explore associations between physical activity and risk of falls and broken or fractured bones in community-dwelling older women. Design, setting and participants: This was a prospective observational survey with 3- and 6-year follow-ups. The sample included 8562 healthy, community-dwelling women, aged 70-75 years in 1996, who completed surveys as participants in the Australian Longitudinal Study on Women’s Health. Outcomes were reports of a fall to the ground, injury from a fall, and broken or fractured bones in 1999 and 2002. The main predictor variable was physical activity level in 1996, categorized based on weekly frequency as none/very low, low, moderate, high, and very high. Covariates were demographic and health-related variables. Logistic regression models were computed separately for each outcome in 1999 and 2002. Main results: In multivariable models, very high physical activity was associated with decreased risk of a fall in 1999 (odds ratio 0.67, 95% CI 0.48 to 0.93) and in 2002 (odds ratio 0.62, 95% CI 0.42 to 0.92). High/very high physical activity was associated with decreased risk of broken or fractured bones in 2002 (odds ratio 0.64, 95% CI 0.42 to 0.96). No significant association was found between physical activity and injury from a fall. Conclusions: The results suggest that at least daily moderate to vigorous physical activity is required for the primary prevention of falls to the ground and broken or fractured bones in women aged 70-75 years.

Physical activity, sitting and depression during pregnancy

Aims The aim of this cross sectional study is to explore levels of physical activity and sitting behaviour amongst a sample of pregnant Australian women (n = 81), and investigate whether reported levels of physical activity and/or time spent sitting were associated with depressive symptom scores after controlling for potential covariates. Methods Study participants were women who attended the antenatal clinic of a large Brisbane maternity hospital between October and November 2006. Data relating to participants. current levels of physical activity, sitting behaviour, depressive symptoms, demographic characteristics and exposure to known risk factors for depression during pregnancy were collected; via on-site survey, follow-up telephone interview (approximately one week later) and post delivery access to participant hospital records. Results Participants were aged 29.5 (¡¾ 5.6) years and mostly partnered (86.4%) with a gross household income above $26,000 per annum (88.9%). Levels of physical activity were generally low, with only 28.4 % of participants reporting sufficient total activity and 16% of participants reporting sufficient planned (leisure-time) activity. The sample mean for depressive symptom scores measured by the Hospital Anxiety and Depression Scale (HADS-D) was 6.38 (¡¾ 2.55). The mean depressive symptom scores for participants who reported total moderate-to-vigorous activity levels of sufficient, insufficient, and none, were 5.43 (¡¾ 1.56), 5.82 (¡¾ 1.77) and 7.63 (¡¾ 3.25), respectively. Hierarchical multivariable linear regression modelling indicated that after controlling for covariates, a statistically significant difference of 1.09 points was observed between mean depressive symptom scores of participants who reported sufficient total physical activity, compared with participants who reported they were engaging in no moderate-to-vigorous activity in a typical week (p = 0.05) but this did not reach the criteria for a clinically meaningful difference. Total physical activity was contributed 2.2% to the total 30.3% of explained variance within this model. The other main contributors to explained variance in multivariable regression models were anxiety symptom scores and the number of existing children. Further, a trend was observed between higher levels of planned sitting behaviour and higher depressive symptom scores (p = 0.06); this correlation was not clinically meaningful. Planned sitting contributed 3.2% to the total 31.3 % of explained variance. The number of regression covariates and limited sample size led to a less than ideal ratio of covariates to participants, probably attenuating this relationship. Specific information about the sitting-based activities in which participants engaged may have provided greater insight about the relationship between planned sitting and depressive symptoms, but these data were not captured by the present study. Conclusions The finding that higher levels of physical activity were associated with lower levels of depressive symptoms is consistent with the current body of existing literature in pregnant women, and with a larger body of evidence based in general population samples. Although this result was not considered clinically meaningful, the criterion for a clinically meaningful result was an a priori decision based on quality of life literature in non-pregnant populations and may not truly reflect a difference in symptoms that is meaningful to pregnant women. Further investigation to establish clinically meaningful criteria for continuous depressive symptom data in pregnant women is required. This result may have implications relating to prevention and management options for depression during pregnancy. The observed trend between planned sitting and depressive symptom scores is consistent with literature based on leisure-time sitting behaviour in general population samples, and suggests that further research in this area, with larger samples of pregnant women and more specific sitting data is required to explore potential associations between activities such as television viewing and depressive symptoms, as this may be an area of behaviour that is amenable to modification.

Work-Related Subjective Experiences, Work-Related Self-Efficacy, and Career Learning Among People with Psychiatric Disabilities

Work-related subjective experiences and work-related self-efficacy were investigated as candidate correlates of career learning among people with schizophrenia and schizoaffective disorder. Work-related self-efficacy was expected to mediate any observed relationship between work-related subjective experiences and employment status, after controlling for demographic, vocational, and clinical covariates. Baseline measures (n 1 = 104) were repeated at six months (n 2 = 94) and 12 months (n 3 = 94). Work-related subjective experiences and work-related self-efficacy were consistently associated with current employment after controlling for covariates. The proposed mediator role of work-related self-efficacy remains a viable hypothesis requiring further investigation. Both work-related subjective experiences and work-related self-efficacy appear promising as components of the social cognitive career learning theory to help explain career development among people with psychiatric disabilities.

A geographical analysis of the role of water supply and sanitation in the risk of helminth infections of children in West Africa

Background We investigated the geographical variation of water supply and sanitation indicators (WS&S) and their role to the risk of schistosomiasis and hookworm infection in school age children in West Africa. The aim was to predict large-scale geographical variation in WS&S, quantify the attributable risk of S. haematobium, S. mansoni and hookworm infections due to WS&S and identify communities where sustainable transmission control could be targeted across the region. Methods National cross-sectional household-based demographic health surveys were conducted in 24,542 households in Burkina Faso, Ghana and Mali, in 2003–2006. We generated spatially-explicit predictions of areas without piped water, toilet facilities and finished floors in West Africa, adjusting for household covariates. Using recently published helminth prevalence data we developed Bayesian geostatistical models (MGB) of S. haematobium, S. mansoni and hookworm infection in West Africa including environmental and the mapped outputs for WS&S. Using these models we estimated the effect of WS&S on parasite risk, quantified their attributable fraction of infection, and mapped the risk of infection in West Africa. Findings Our maps show that most areas in West Africa are very poorly served by water supply except in major urban centers. There is a better geographical coverage for toilet availability and improved household flooring. We estimated smaller attributable risks for water supply in S. mansoni (47%) compared to S. haematobium (71%), and 5% of hookworm cases could be averted by improving sanitation. Greater levels of inadequate sanitation increased the risk of schistosomiasis, and increased levels of unsafe water supply increased the risk of hookworm. The role of floor type for S. haematobium infection (21%) was comparable to that of S. mansoni (16%), but was significantly higher for hookworm infection (86%). S. haematobium and hookworm maps accounting for WS&S show small clusters of maximal prevalence areas in areas bordering Burkina Faso and Mali smaller. The map of S. mansoni shows that this parasite is much more wide spread across the north of the Niger River basin than previously predicted. Interpretation Our maps identify areas where the Millennium Development Goal for water and sanitation is lagging behind. Our results show that WS&S are important contributors to the burden of major helminth infections of children in West Africa. Including information about WS&S as well as the “traditional” environmental risk factors in spatial models of helminth risk yielded a substantial gain both in model fit and at explaining the proportion of spatial variance in helminth risk. Mapping the distribution of infection risk adjusted for WS&S allowed the identification of communities in West Africa where integrative preventive chemotherapy and engineering interventions will yield the greatest public health benefits.