Recent changes to US research funding are having far-reaching consequences that imperil the integrity of science and the provision of care to vulnerable populations. Resisting these changes, the BJPsych Portfolio reaffirms its commitment to publishing mental science and advancing psychiatric knowledge that improves the mental health of one and all.

When using machine learning to model environmental systems, it is often a model’s ability to predict extreme behaviors that yields the highest practical value to policy makers. However, most existing error metrics used to evaluate the performance of environmental machine learning models weigh error equally across test data. Thus, routine performance is prioritized over a model’s ability to robustly quantify extreme behaviors. In this work, we present a new error metric, termed Reflective Error, which quantifies the degree at which our model error is distributed around our extremes, in contrast to existing model evaluation methods that aggregate error over all events. The suitability of our proposed metric is demonstrated on a real-world hydrological modeling problem, where extreme values are of particular concern.

Machine learning models have been used extensively in hydrology, but issues persist with regard to their transparency, and there is currently no identifiable best practice for forcing variables in streamflow or flood modeling. In this paper, using data from the Centre for Ecology & Hydrology’s National River Flow Archive and from the European Centre for Medium-Range Weather Forecasts, we present a study that focuses on the input variable set for a neural network streamflow model to demonstrate how certain variables can be internalized, leading to a compressed feature set. By highlighting this capability to learn effectively using proxy variables, we demonstrate a more transferable framework that minimizes sensing requirements and that enables a route toward generalizing models.

Approximately one-third of individuals in a major depressive episode will not achieve sustained remission despite multiple, well-delivered treatments. These patients experience prolonged suffering and disproportionately utilize mental and general health care resources. The recently proposed clinical heuristic of ‘difficult-to-treat depression’ (DTD) aims to broaden our understanding and focus attention on the identification, clinical management, treatment selection, and outcomes of such individuals. Clinical trial methodologies developed to detect short-term therapeutic effects in treatment-responsive populations may not be appropriate in DTD. This report reviews three essential challenges for clinical intervention research in DTD: (1) how to define and subtype this heterogeneous group of patients; (2) how, when, and by what methods to select, acquire, compile, and interpret clinically meaningful outcome metrics; and (3) how to choose among alternative clinical trial design options to promote causal inference and generalizability. The boundaries of DTD are uncertain, and an evidence-based taxonomy and reliable assessment tools are preconditions for clinical research and subtyping. Traditional outcome metrics in treatment-responsive depression may not apply to DTD, as they largely reflect the only short-term symptomatic change and do not incorporate durability of benefit, side effect burden, or sustained impact on quality of life or daily function. The trial methodology will also require modification as trials will likely be of longer duration to examine the sustained impact, raising complex issues regarding control group selection, blinding and its integrity, and concomitant treatments.

Strong associations between neural tube defects (NTDs) and monozygotic (MZ) twinning have long been noted, and it has been suggested that NTD cases who do not present as MZ twins may be the survivors of MZ twinning events. We have recently shown that MZ twins carry a strong, distinctive DNA methylation signature and have developed an algorithm based on genomewide DNA methylation array data that distinguishes MZ twins from dizygotic twins and other relatives at well above chance level. We have applied this algorithm to published methylation data from five fetal tissues (placental chorionic villi, kidney, spinal cord, brain and muscle) collected from spina bifida cases (n = 22), anencephalic cases (n = 15) and controls (n = 19). We see no difference in signature between cases and controls, providing no support for a common etiological role of MZ twinning in NTDs. The strong associations therefore continue to await elucidation.

In 2008, we saw an emerging business need to accurately identify and develop individuals early in their careers who have the potential to be effective leaders in later organizational positions. We decided early on to take a comprehensive and systematic approach to the challenge in order to build effective solutions with real organizational impact that are sustainable over time. We knew that this was a complex problem and that there was little agreement in industrial–organizational (I-O) psychology on how to approach this critical business need and what approaches would be most effective. Since then, we moved through the following process stages to arrive at effective solutions that are now being used in numerous organizations.

The past four decades have witnessed a transformation in research on thebenefits of psychological therapies. However, even though therapistshighlight that negative and adverse effects are seen in day-to-day practice,research on the negative effects of psychotherapy is insufficient. Given theunrelenting popularity of therapies, the argument for examining the adverseeffects of psychotherapy would seem to be compelling. Such a strategy wouldextend beyond supervision of individual therapists to the introduction ofmonitoring systems that allow for a more systematic examination of failedpsychotherapy interventions (such as exist for medication prescribing). Thestarting point could be the development of a consensus on how to define,classify and assess psychotherapy side-effects, unwanted events, adversereactions, etc. This would provide a conceptual framework for communication,monitoring and research. This approach should not be viewed as an attack ontherapies: every branch of medicine learns from mistakes, the same mustsurely be true for psychological treatments.

Few studies have investigated developmental strengths and weaknesses within the cognitive profile of children and adolescents with fragile X syndrome (FXS), a single-gene cause of inherited intellectual impairment. With a prospective longitudinal design and using normalized raw scores (Z scores) to circumvent floor effects, we measured cognitive functioning of 184 children and adolescents with FXS (ages 6 to 16) using the Wechsler Scale of Intelligence for Children on one to three occasions for each participant. Participants with FXS received lower raw scores relative to the Wechsler Scale of Intelligence for Children normative sample across the developmental period. Verbal comprehension, perceptual organization, and processing speed Z scores were marked by a widening gap from the normative sample, while freedom from distractibility Z scores showed a narrowing gap. Key findings include a relative strength for verbal skills in comparison with visuospatial–constructive skills arising in adolescence and a discrepancy between working memory (weakness) and processing speed (strength) in childhood that diminishes in adolescence. Results suggest that the cognitive profile associated with FXS develops dynamically from childhood to adolescence. Findings are discussed within the context of aberrant brain morphology in childhood and maturation in adolescence. We argue that assessing disorder-specific cognitive developmental profiles will benefit future disorder-specific treatment research.

Despite more than a decade of heightened defence spending and active fighting in the War in Afghanistan, the longest combat operation in the history of the Canadian Forces, scholars know precious little about how the socio-demographic characteristics and attitudes of Canadians may influence their views about taking part in overseas combat operations and funding the institution charged with carrying out these dangerous activities. By testing a range of hypotheses, which purport to explain the influence of multiple socio-demographic and attitudinal factors on Canadians' attitudes toward defence spending and the participation of the Canadian Forces in overseas combat operations, against data from the 2004 and 2011 Canadian Election Study, this article ascertains the most important determinants of Canadians' preferences about defence spending and the use of military force by the Government of Canada.

The genetic basis of cardiovascular disease (CVD) is complex and still largely elusive. Plasma lipid concentrations are well-established risk factors for cardiovascular disease (CVD), and have adult heritabilities ranging from 0.48 to 0.87. Estimates for adolescents are slightly higher (range 0.71 to 0.82). To identify loci affecting lipid concentrations across adolescence, we analyzed longitudinal lipid data in a sample of 134 monozygotic and 626 dizygotic twin pairs at ages twelve, fourteen and sixteen, and their siblings, from 760 Australian families. Univariate linkage analysis for each phenotype and time point was supplemented by multivariate analysis across the time points. A genome-wide association scan was also performed on a subset of the subjects (N = 441). The strongest linkage was seen for triglycerides on chromosome 6p24.3 (multivariate –log10p = 6.81; equivalent LOD = 6.13; p = 1.55 × 10–7). Significant linkage was also found for LDL cholesterol on chromosome 2q35 (multivariate –log10p = 5.59; equivalent LOD = 4.53; p = 2.57 × 10–6). In the association analysis, rs10503840 on 8p21.1 was significantly associated with total cholesterol levels at age fourteen (p = 8.24 × 10–7, estimated significance threshold 2.45 × 10–6). Association at p < 2.25 × 10–6 was also found between triglycerides at age 12 and rs10507266, in an intron of THRAP2 (MIM 608771) on 12q24.21; and between HDL-C at age 14 and rs10506325 in an intergenic region of 12q13.13. Suggestive evidence of association at ages twelve and fourteen was found between HDL-C and rs10492859 on 16q23 (p = 2.42 × 10–5 and 2.77 × 10–4, respectively). Further longitudinal genetic studies of cardiovascular risk factors, focused on critical periods of development or change, are needed.