The impact of spatial scales and spatial smoothing on the outcome of Bayesian spatial model

Discretization of a geographical region is quite common in spatial analysis. There have been few studies into the impact of different geographical scales on the outcome of spatial models for different spatial patterns. This study aims to investigate the impact of spatial scales and spatial smoothing on the outcomes of modelling spatial point-based data. Given a spatial point-based dataset (such as occurrence of a disease), we study the geographical variation of residual disease risk using regular grid cells. The individual disease risk is modelled using a logistic model with the inclusion of spatially unstructured and/or spatially structured random effects. Three spatial smoothness priors for the spatially structured component are employed in modelling, namely an intrinsic Gaussian Markov random field, a second-order random walk on a lattice, and a Gaussian field with Matern correlation function. We investigate how changes in grid cell size affect model outcomes under different spatial structures and different smoothness priors for the spatial component. A realistic example (the Humberside data) is analyzed and a simulation study is described. Bayesian computation is carried out using an integrated nested Laplace approximation. The results suggest that the performance and predictive capacity of the spatial models improve as the grid cell size decreases for certain spatial structures. It also appears that different spatial smoothness priors should be applied for different patterns of point data.

Incidence of unilateral arm lymphoedema after breast cancer : a systematic review and meta-analysis

Background The body of evidence related to breast-cancer-related lymphoedema incidence and risk factors has substantially grown and improved in quality over the past decade. We assessed the incidence of unilateral arm lymphoedema after breast cancer and explored the evidence available for lymphoedema risk factors. Methods We searched Academic Search Elite, Cumulative Index to Nursing and Allied Health, Cochrane Central Register of Controlled Trials (clinical trials), and Medline for research articles that assessed the incidence or prevalence of, or risk factors for, arm lymphoedema after breast cancer, published between January 1, 2000, and June 30, 2012. We extracted incidence data and calculated corresponding exact binomial 95% CIs. We used random effects models to calculate a pooled overall estimate of lymphoedema incidence, with subgroup analyses to assess the effect of different study designs, countries of study origin, diagnostic methods, time since diagnosis, and extent of axillary surgery. We assessed risk factors and collated them into four levels of evidence, depending on consistency of findings and quality and quantity of studies contributing to findings. Findings 72 studies met the inclusion criteria for the assessment of lymphoedema incidence, giving a pooled estimate of 16·6% (95% CI 13·6–20·2). Our estimate was 21·4% (14·9–29·8) when restricted to data from prospective cohort studies (30 studies). The incidence of arm lymphoedema seemed to increase up to 2 years after diagnosis or surgery of breast cancer (24 studies with time since diagnosis or surgery of 12 to <24 months; 18·9%, 14·2–24·7), was highest when assessed by more than one diagnostic method (nine studies; 28·2%, 11·8–53·5), and was about four times higher in women who had an axillary-lymph-node dissection (18 studies; 19·9%, 13·5–28·2) than it was in those who had sentinel-node biopsy (18 studies; 5·6%, 6·1–7·9). 29 studies met the inclusion criteria for the assessment of risk factors. Risk factors that had a strong level of evidence were extensive surgery (ie, axillary-lymph-node dissection, greater number of lymph nodes dissected, mastectomy) and being overweight or obese. Interpretation Our findings suggest that more than one in five women who survive breast cancer will develop arm lymphoedema. A clear need exists for improved understanding of contributing risk factors, as well as of prevention and management strategies to reduce the individual and public health burden of this disabling and distressing disorder.

Combining expert opinions in prior elicitation

We consider the problem of combining opinions from different experts in an explicitly model-based way to construct a valid subjective prior in a Bayesian statistical approach. We propose a generic approach by considering a hierarchical model accounting for various sources of variation as well as accounting for potential dependence between experts. We apply this approach to two problems. The first problem deals with a food risk assessment problem involving modelling dose-response for Listeria monocytogenes contamination of mice. Two hierarchical levels of variation are considered (between and within experts) with a complex mathematical situation due to the use of an indirect probit regression. The second concerns the time taken by PhD students to submit their thesis in a particular school. It illustrates a complex situation where three hierarchical levels of variation are modelled but with a simpler underlying probability distribution (log-Normal).

Survival models of community tenure and length of hospital stay for the seriously mentally ill : a 10-year perspective

Objective: To examine the effects of personal and community characteristics, specifically race and rurality, on lengths of state psychiatric hospital and community stays using maximum likelihood survival analysis with a special emphasis on change over a ten year period of time. Data Sources: We used the administrative data of the Virginia Department of Mental Health, Mental Retardation, and Substance Abuse Services (DMHMRSAS) from 1982-1991 and the Area Resources File (ARF). Given these two sources, we constructed a history file for each individual who entered the state psychiatric system over the ten year period. Histories included demographic, treatment, and community characteristics. Study Design: We used a longitudinal, population-based design with maximum likelihood estimation of survival models. We presented a random effects model with unobserved heterogeneity that was independent of observed covariates. The key dependent variables were lengths of inpatient stay and subsequent length of community stay. Explanatory variables measured personal, diagnostic, and community characteristics, as well as controls for calendar time. Data Collection: This study used secondary, administrative, and health planning data. Principal Findings: African-American clients leave the community more quickly than whites. After controlling for other characteristics, however, race does not affect hospital length of stay. Rurality does not affect length of community stays once other personal and community characteristics are controlled for. However, people from rural areas have longer hospital stays even after controlling for personal and community characteristics. The effects of time are significantly smaller than expected. Diagnostic composition effects and a decrease in the rate of first inpatient admissions explain part of this reduced impact of time. We also find strong evidence for the existence of unobserved heterogeneity in both types of stays and adjust for this in our final models. Conclusions: Our results show that information on client characteristics available from inpatient stay records is useful in predicting not only the length of inpatient stay but also the length of the subsequent community stay. This information can be used to target increased discharge planning for those at risk of more rapid readmission to inpatient care. Correlation across observed and unobserved factors affecting length of stay has significant effects on the measurement of relationships between individual factors and lengths of stay. Thus, it is important to control for both observed and unobserved factors in estimation.

An investigation of the impact of various geographical scales for the specification of spatial dependence

Ecological studies are based on characteristics of groups of individuals, which are common in various disciplines including epidemiology. It is of great interest for epidemiologists to study the geographical variation of a disease by accounting for the positive spatial dependence between neighbouring areas. However, the choice of scale of the spatial correlation requires much attention. In view of a lack of studies in this area, this study aims to investigate the impact of differing definitions of geographical scales using a multilevel model. We propose a new approach -- the grid-based partitions and compare it with the popular census region approach. Unexplained geographical variation is accounted for via area-specific unstructured random effects and spatially structured random effects specified as an intrinsic conditional autoregressive process. Using grid-based modelling of random effects in contrast to the census region approach, we illustrate conditions where improvements are observed in the estimation of the linear predictor, random effects, parameters, and the identification of the distribution of residual risk and the aggregate risk in a study region. The study has found that grid-based modelling is a valuable approach for spatially sparse data while the SLA-based and grid-based approaches perform equally well for spatially dense data.

Awards before and after the Nobel Prize : a Matthew effect and/or a ticket to one's own funeral?

The primary aim of this descriptive exploration of scientists’ life cycle award patterns is to evaluate whether awards breed further awards and identify researcher experiences after reception of the Nobel Prize. To achieve this goal, we collected data on the number of awards received each year for 50 years before and after Nobel Prize reception by all 1901–2000 Nobel laureates in physics, chemistry, and medicine or physiology. Our results indicate an increasing rate of awards before Nobel reception, reaching the summit precisely in the year of the Nobel Prize. After this pinnacle year, awards drop sharply. This result is confirmed by separate analyses of three different disciplines and by a random-effects negative binomial regression model. Such an effect, however, does not emerge for more recent Nobel laureates (1971–2000). In addition, Nobelists in medicine or physiology generate more awards shortly before and after prize reception, whereas laureates in chemistry attract more awards as time progresses.

Multilevel competing risk models to evaluate the risk of nosocomial infection

Introduction Risk factor analyses for nosocomial infections (NIs) are complex. First, due to competing events for NI, the association between risk factors of NI as measured using hazard rates may not coincide with the association using cumulative probability (risk). Second, patients from the same intensive care unit (ICU) who share the same environmental exposure are likely to be more similar with regard to risk factors predisposing to a NI than patients from different ICUs. We aimed to develop an analytical approach to account for both features and to use it to evaluate associations between patient- and ICU-level characteristics with both rates of NI and competing risks and with the cumulative probability of infection. Methods We considered a multicenter database of 159 intensive care units containing 109,216 admissions (813,739 admission-days) from the Spanish HELICS-ENVIN ICU network. We analyzed the data using two models: an etiologic model (rate based) and a predictive model (risk based). In both models, random effects (shared frailties) were introduced to assess heterogeneity. Death and discharge without NI are treated as competing events for NI. Results There was a large heterogeneity across ICUs in NI hazard rates, which remained after accounting for multilevel risk factors, meaning that there are remaining unobserved ICU-specific factors that influence NI occurrence. Heterogeneity across ICUs in terms of cumulative probability of NI was even more pronounced. Several risk factors had markedly different associations in the rate-based and risk-based models. For some, the associations differed in magnitude. For example, high Acute Physiology and Chronic Health Evaluation II (APACHE II) scores were associated with modest increases in the rate of nosocomial bacteremia, but large increases in the risk. Others differed in sign, for example respiratory vs cardiovascular diagnostic categories were associated with a reduced rate of nosocomial bacteremia, but an increased risk. Conclusions A combination of competing risks and multilevel models is required to understand direct and indirect risk factors for NI and distinguish patient-level from ICU-level factors.

Exploring association between perceived importance of travel/traffic information and travel behaviour in natural disasters : a case study of the 2011 Brisbane floods

A sound understanding of travellers’ behavioural changes and adaptation when facing a natural disaster is a key factor in efficiently and effectively managing transport networks at such times. This study specifically investigates the importance of travel/traffic information and its impact on travel behaviour during natural disasters. Using the 2011 Brisbane flood as a case study, survey respondents’ perceptions of the importance of travel/traffic information before, during, and after the flood were modelled using random-effects ordered logit. A hysteresis phenomenon was observed: respondents’ perceptions of the importance of travel/traffic information increased during the flood, and although its perceived importance decreased after the flood, it did not return to the pre-flood level. Results also reveal that socio-demographic features (such as gender and age) have a significant impact on respondents’ perceptions of the importance of travel/traffic information. The roles of travel time and safety in a respondent’s trip planning are also significantly correlated to their perception of the importance of this information. The analysis further shows that during the flood, respondents generally thought that travel/traffic information was important, and adjusted their travel plans according to information received. When controlling for other factors, the estimated odds of changing routes and cancelling trips for a respondent who thought that travel/traffic information was important, are respectively about three times and seven times the estimated odds for a respondent who thought that travel/traffic information was not important. In contrast, after the flood, the influence of travel/traffic information on respondents’ travel behaviour diminishes. Finally, the analysis shows no evidence of the influence of travel/traffic information’s on respondents’ travel mode; this indicates that inducing travel mode change is a challenging task.

What roles do contemporaneous and cumulative incomes play in the income–child health gradient for young children? Evidence from an Australian Panel

The literature to date shows that children from poorer households tend to have worse health than their peers, and the gap between them grows with age. We investigate whether and how health shocks (as measured by the onset of chronic conditions) contribute to the income–child health gradient and whether the contemporaneous or cumulative effects of income play important mitigating roles. We exploit a rich panel dataset with three panel waves called the Longitudinal Study of Australian children. Given the availability of three waves of data, we are able to apply a range of econometric techniques (e.g. fixed and random effects) to control for unobserved heterogeneity. The paper makes several contributions to the extant literature. First, it shows that an apparent income gradient becomes relatively attenuated in our dataset when the cumulative and contemporaneous effects of household income are distinguished econometrically. Second, it demonstrates that the income–child health gradient becomes statistically insignificant when controlling for parental health and health-related behaviours or unobserved heterogeneity.

Genetic variants underlying risk of endometriosis: Insights from meta-analysis of eight genome-wide association and replication datasets

BACKGROUND Endometriosis is a heritable common gynaecological condition influenced by multiple genetic and environmental factors. Genome-wide association studies (GWASs) have proved successful in identifying common genetic variants of moderate effects for various complex diseases. To date, eight GWAS and replication studies from multiple populations have been published on endometriosis. In this review, we investigate the consistency and heterogeneity of the results across all the studies and their implications for an improved understanding of the aetiology of the condition. METHODS Meta-analyses were conducted on four GWASs and four replication studies including a total of 11 506 cases and 32 678 controls, and on the subset of studies that investigated associations for revised American Fertility Society (rAFS) Stage III/IV including 2859 cases. The datasets included 9039 cases and 27 343 controls of European (Australia, Belgium, Italy, UK, USA) and 2467 cases and 5335 controls of Japanese ancestry. Fixed and Han and Elkin random-effects models, and heterogeneity statistics (Cochran's Q test), were used to investigate the evidence of the nine reported genome-wide significant loci across datasets and populations. RESULTS Meta-analysis showed that seven out of nine loci had consistent directions of effect across studies and populations, and six out of nine remained genome-wide significant (P < 5 × 10(-8)), including rs12700667 on 7p15.2 (P = 1.6 × 10(-9)), rs7521902 near WNT4 (P = 1.8 × 10(-15)), rs10859871 near VEZT (P = 4.7 × 10(-15)), rs1537377 near CDKN2B-AS1 (P = 1.5 × 10(-8)), rs7739264 near ID4 (P = 6.2 × 10(-10)) and rs13394619 in GREB1 (P = 4.5 × 10(-8)). In addition to the six loci, two showed borderline genome-wide significant associations with Stage III/IV endometriosis, including rs1250248 in FN1 (P = 8 × 10(-8)) and rs4141819 on 2p14 (P = 9.2 × 10(-8)). Two independent inter-genic loci, rs4141819 and rs6734792 on chromosome 2, showed significant evidence of heterogeneity across datasets (P < 0.005). Eight of the nine loci had stronger effect sizes among Stage III/IV cases, implying that they are likely to be implicated in the development of moderate to severe, or ovarian, disease. While three out of nine loci were inter-genic, the remaining were in or near genes with known functions of biological relevance to endometriosis, varying from roles in developmental pathways to cellular growth/carcinogenesis. CONCLUSIONS Our meta-analysis shows remarkable consistency in endometriosis GWAS results across studies, with little evidence of population-based heterogeneity. They also show that the phenotypic classifications used in GWAS to date have been limited. Stronger associations with Stage III/IV disease observed for most loci emphasize the importance for future studies to include detailed sub-phenotype information. Functional studies in relevant tissues are needed to understand the effect of the variants on downstream biological pathways.

Brain fiber architecture, genetics, and intelligence: a high angular resolution diffusion imaging (HARDI) study

We developed an analysis pipeline enabling population studies of HARDI data, and applied it to map genetic influences on fiber architecture in 90 twin subjects. We applied tensor-driven 3D fluid registration to HARDI, resampling the spherical fiber orientation distribution functions (ODFs) in appropriate Riemannian manifolds, after ODF regularization and sharpening. Fitting structural equation models (SEM) from quantitative genetics, we evaluated genetic influences on the Jensen-Shannon divergence (JSD), a novel measure of fiber spatial coherence, and on the generalized fiber anisotropy (GFA) a measure of fiber integrity. With random-effects regression, we mapped regions where diffusion profiles were highly correlated with subjects' intelligence quotient (IQ). Fiber complexity was predominantly under genetic control, and higher in more highly anisotropic regions; the proportion of genetic versus environmental control varied spatially. Our methods show promise for discovering genes affecting fiber connectivity in the brain.

Brain network efficiency and topology depend on the fiber tracking method: 11 tractography algorithms compared in 536 subjects

As connectivity analyses become more popular, claims are often made about how the brain's anatomical networks depend on age, sex, or disease. It is unclear how results depend on tractography methods used to compute fiber networks. We applied 11 tractography methods to high angular resolution diffusion images of the brain (4-Tesla 105-gradient HARDI) from 536 healthy young adults. We parcellated 70 cortical regions, yielding 70×70 connectivity matrices, encoding fiber density. We computed popular graph theory metrics, including network efficiency, and characteristic path lengths. Both metrics were robust to the number of spherical harmonics used to model diffusion (4th-8th order). Age effects were detected only for networks computed with the probabilistic Hough transform method, which excludes smaller fibers. Sex and total brain volume affected networks measured with deterministic, tensor-based fiber tracking but not with the Hough method. Each tractography method includes different fibers, which affects inferences made about the reconstructed networks.

Whole-body cryotherapy (extreme cold air exposure) for preventing and treating muscle soreness after exercise in adults

Background Recovery strategies are often usedwith the intention of preventing orminimisingmuscle soreness after exercise. Whole-body cryotherapy, which involves a single or repeated exposure(s) to extremely cold dry air (below -100 °C) in a specialised chamber or cabin for two to four minutes per exposure, is currently being advocated as an effective intervention to reduce muscle soreness after exercise. Objectives To assess the effects (benefits and harms) of whole-body cryotherapy (extreme cold air exposure) for preventing and treating muscle soreness after exercise in adults. Search methods We searched the Cochrane Bone, Joint and Muscle Trauma Group Specialised Register, the Cochrane Central Register of Controlled Trials, MEDLINE, EMBASE, CINAHL, the British Nursing Index and the Physiotherapy Evidence Database. We also searched the reference lists of articles, trial registers and conference proceedings, handsearched journals and contacted experts. The searches were run in August 2015. Selection criteria We aimed to include randomised and quasi-randomised trials that compared the use of whole-body cryotherapy (WBC) versus a passive or control intervention (rest, no treatment or placebo treatment) or active interventions including cold or contrast water immersion, active recovery and infrared therapy for preventing or treating muscle soreness after exercise in adults. We also aimed to include randomised trials that compared different durations or dosages of WBC. Our prespecified primary outcomes were muscle soreness, subjective recovery (e.g. tiredness, well-being) and adverse effects. Data collection and analysis Two review authors independently screened search results, selected studies, assessed risk of bias and extracted and cross-checked data. Where appropriate, we pooled results of comparable trials. The random-effects model was used for pooling where there was substantial heterogeneity.We assessed the quality of the evidence using GRADE. Main results Four laboratory-based randomised controlled trials were included. These reported results for 64 physically active predominantly young adults (mean age 23 years). All but four participants were male. Two trials were parallel group trials (44 participants) and two were cross-over trials (20 participants). The trials were heterogeneous, including the type, temperature, duration and frequency of WBC, and the type of preceding exercise. None of the trials reported active surveillance of predefined adverse events. All four trials had design features that carried a high risk of bias, potentially limiting the reliability of their findings. The evidence for all outcomes was classified as ’very low’ quality based on the GRADE criteria. Two comparisons were tested: WBC versus control (rest or no WBC), tested in four studies; and WBC versus far-infrared therapy, also tested in one study. No studies compared WBC with other active interventions, such as cold water immersion, or different types and applications of WBC. All four trials compared WBC with rest or no WBC. There was very low quality evidence for lower self-reported muscle soreness (pain at rest) scores after WBC at 1 hour (standardised mean difference (SMD) -0.77, 95% confidence interval (CI) -1.42 to -0.12; 20 participants, 2 cross-over trials); 24 hours (SMD -0.57, 95%CI -1.48 to 0.33) and 48 hours (SMD -0.58, 95% CI -1.37 to 0.21), both with 38 participants, 2 cross-over studies, 1 parallel group study; and 72 hours (SMD -0.65, 95% CI -2.54 to 1.24; 29 participants, 1 cross-over study, 1 parallel group study). Of note is that the 95% CIs also included either no between-group differences or a benefit in favour of the control group. One small cross-over trial (9 participants) found no difference in tiredness but better well-being after WBC at 24 hours post exercise. There was no report of adverse events. One small cross-over trial involving nine well-trained runners provided very low quality evidence of lower levels of muscle soreness after WBC, when compared with infrared therapy, at 1 hour follow-up, but not at 24 or 48 hours. The same trial found no difference in well-being but less tiredness after WBC at 24 hours post exercise. There was no report of adverse events. Authors’ conclusions There is insufficient evidence to determine whether whole-body cryotherapy (WBC) reduces self-reportedmuscle soreness, or improves subjective recovery, after exercise compared with passive rest or no WBC in physically active young adult males. There is no evidence on the use of this intervention in females or elite athletes. The lack of evidence on adverse events is important given that the exposure to extreme temperature presents a potential hazard. Further high-quality, well-reported research in this area is required and must provide detailed reporting of adverse events.

Prospective meta-analysis of interleukin 1 gene complex polymorphisms confirms associations with ankylosing spondylitis

Objectives: The aim of the current study was to determine the contribution of interleukin (IL) 1 gene cluster polymorphisms previously implicated in susceptibility for ankylosing spondylitis (AS) to AS susceptibility in different populations worldwide. Methods: Nine polymorphisms in the IL1 gene cluster members IL1A (rs2856836, rs17561 and rs1894399), IL1B (rs16944), IL1F10 (rs3811058) and IL1RN (rs419598, the IL1RA VNTR, rs315952 and rs315951) were genotyped in 2675 AS cases and 2592 healthy controls recruited in 12 different centres in 10 countries. Association of variants with AS was tested by Mantel-Haenszel random effects analysis. Results: Strong association was observed with three single nucleotide polymorphisms (SNPs) in the IL1A gene (rs2856836, rs17561, rs1894399, p = 0.0036, 0.000019 and 0.0003, respectively). There was no evidence of significant heterogeneity of effects between centres, and no evidence of non-combinability of findings. The population attributable risk fraction of these variants in Caucasians is estimated at 4-6%. Conclusions: This study confirms that IL1A is associated with susceptibility to AS. Association of the other IL1 gene complex members could not be excluded in specific populations. Prospective meta-analysis is a useful tool in confirmation studies of genes associated with complex genetic disorders such as AS, providing sufficiently large sample sizes to produce robust findings often not achieved in smaller individual cohorts.

Genome-wide association study for radiographic vertebral fractures: A potential role for the 16q24 BMD locus

Vertebral fracture risk is a heritable complex trait. The aim of this study was to identify genetic susceptibility factors for osteoporotic vertebral fractures applying a genome-wide association study (GWAS) approach. The GWAS discovery was based on the Rotterdam Study, a population-based study of elderly Dutch individuals aged >55years; and comprising 329 cases and 2666 controls with radiographic scoring (McCloskey-Kanis) and genetic data. Replication of one top-associated SNP was pursued by de-novo genotyping of 15 independent studies across Europe, the United States, and Australia and one Asian study. Radiographic vertebral fracture assessment was performed using McCloskey-Kanis or Genant semi-quantitative definitions. SNPs were analyzed in relation to vertebral fracture using logistic regression models corrected for age and sex. Fixed effects inverse variance and Han-Eskin alternative random effects meta-analyses were applied. Genome-wide significance was set at p<5×10-8. In the discovery, a SNP (rs11645938) on chromosome 16q24 was associated with the risk for vertebral fractures at p=4.6×10-8. However, the association was not significant across 5720 cases and 21,791 controls from 14 studies. Fixed-effects meta-analysis summary estimate was 1.06 (95% CI: 0.98-1.14; p=0.17), displaying high degree of heterogeneity (I2=57%; Qhet p=0.0006). Under Han-Eskin alternative random effects model the summary effect was significant (p=0.0005). The SNP maps to a region previously found associated with lumbar spine bone mineral density (LS-BMD) in two large meta-analyses from the GEFOS consortium. A false positive association in the GWAS discovery cannot be excluded, yet, the low-powered setting of the discovery and replication settings (appropriate to identify risk effect size >1.25) may still be consistent with an effect size <1.10, more of the type expected in complex traits. Larger effort in studies with standardized phenotype definitions is needed to confirm or reject the involvement of this locus on the risk for vertebral fractures.