This study investigated the relationship between hospital exposures, intestinal microbiota, and subsequent risk of Clostridium difficile-associated disease (CDAD), with use of a nested case-control design. The study included 599 patients, hospitalized from September 2006 through May 2007 in Montreal, Quebec, from whom fecal samples were obtained within 72 h after admission; 25 developed CDAD, and 50 matched controls were selected for analysis. Nonsteroidal anti-inflammatory drugs and antibiotic use were associated with CDAD. Fecal specimens were evaluated by 16S ribosomal RNA microarray to characterize bacteria in the intestinal microbiota during the at-risk period. Probe intensities were higher for Firmicutes, Proteobacteria, and Actinobacteria in the patients with CDAD, compared with controls, whereas probe intensities for Bacteroidetes were lower. After epidemiologic factors were controlled for, only Bacteroidetes and Firmicutes remained significantly and independently associated with development of CDAD. Hospital exposures were associated with changes in the intestinal microbiota and risk of CDAD, and these changes were not driven exclusively by antimicrobial use.

Objective Subarachnoid hemorrhage (SAH) is associated with inflammation which may mediate poor outcome in SAH. We hypothesize that elevated serum tumor necrosis factor-alpha (TNFα) and interleukin-6 (IL-6) are associated with vasospasm and poor outcome in SAH. Methods In 52 consecutive SAH subjects, we compared TNFα and IL-6 levels on post-SAH days 0–1, 2–3, 4–5, 6–8, and 10–14 with respect to vasospasm and to poor outcome at 3- and 6-months. Vasospasm was defined as >50% reduction in vessel caliber on angiography. Poor outcome was defined as modified Rankin score >2. Results Elevated TNFα on post-SAH days 2–3 was associated with poor 3-month outcome (p=0.0004). Global elevation of TNFα over time (post-SAH days 0–14) was independently associated with poor 3-month outcome after adjusting for Hunt-and-Hess grade and age (p=0.02). Neither cross-sectional nor IL-6 levels over time were associated with outcome. Neither TNFα nor IL-6 levels were associated with vasospasm. Conclusions Elevation in serum TNFα on post-SAH days 2–3 and global elevation of TNFα over time are associated with poor outcome but not with angiographic vasospasm in this small cohort. Future studies are needed to define the role of TNFα in SAH-related brain injury and its potential as a SAH outcome biomarker.

Attention deficit/hyperactivity disorder (ADHD) is a heterogeneous disorder characterized by inappropriate levels of attention, hyperactivity, and impulsivity. Although a strong genetic component to the disorder has been established, the molecular genetic underpinnings of this disorder remain elusive. Recently, several studies have reported an association between polymorphisms within the latrophilin 3 gene (LPHN3) and ADHD. Interestingly, the same single-nucleotide polymorphism conferring susceptibility to ADHD has also been found to predict efficacy of stimulant medication in children. The main objectives of the current article are: (i) To tackle the phenotype heterogeneity issue in ADHD by defining an objective and quantitative measure of response to treatment in a sample of ADHD children based on a hand held automatic device (Actiwatch) and (ii) to use this measure to reproduce for the first time the association between LPHN3 variants and response to methylphenidate (MPH) using a double-blind, placebo-controlled crossover experimental design. The results of our study confirm the hypothesis that LPHN3 is associated with response to MPH in ADHD children. Although this will require further validation, our work suggests that the use of an objective measure of response to treatment, such as the change in the child's motor activity measured by Actiwatch, has the potential to uncover genetic association signals that in some conditions might not be obtained using more subjective measures, such as the clinical consensus rating, for example.

Estimating causal exposure effects in observational studies ideally requires the analyst to have a vast knowledge of the domain of application. Investigators often bypass difficulties related to the identification and selection of confounders through the use of fully adjusted outcome regression models. However, since such models likely contain more covariates than required, the variance of the regression coefficient for exposure may be unnecessarily large. Instead of using a fully adjusted model, model selection can be attempted. Most classical statistical model selection approaches, such as Bayesian model averaging, do not readily address causal effect estimation. We present a new model averaged approach to causal inference, Bayesian causal effect estimation (BCEE), which is motivated by the graphical framework for causal inference. BCEE aims to unbiasedly estimate the causal effect of a continuous exposure on a continuous outcome while being more efficient than a fully adjusted approach.