Explore the vast range of titles available.

Psychiatric research has entered the age of ‘Big Data’. Datasets now routinely involve thousands of heterogeneous variables, including clinical, neuroimaging, genomic, proteomic, transcriptomic and other ‘omic’ measures. The analysis of these datasets is challenging, especially when the number of measurements exceeds the number of individuals, and may be further complicated by missing data for some subjects and variables that are highly correlated. Statistical learning-based models are a natural extension of classical statistical approaches but provide more effective methods to analyse very large datasets. In addition, the predictive capability of such models promises to be useful in developing decision support systems. That is, methods that can be introduced to clinical settings and guide, for example, diagnosis classification or personalized treatment. In this review, we aim to outline the potential benefits of statistical learning methods in clinical research. We first introduce the concept of Big Data in different environments. We then describe how modern statistical learning models can be used in practice on Big Datasets to extract relevant information. Finally, we discuss the strengths of using statistical learning in psychiatric studies, from both research and practical clinical points of view.

An updated review of 34 human observational studies indicates that the length polymorphism of the serotonin transporter gene moderates the effect of environmental adversity in the development of depression. This finding depends on the use of contextual or objective methods to assess environmental adversity and is attenuated when self-report instruments are used. Inconsistent findings in male adolescents suggest a developmental stage and sex-specific protective mechanism. These systematic relationships between method and results should be followed up to specify causal mechanisms leading to depression.

Gene–environmental interaction (G × E) between a common functional polymorphism in the promoter region of the serotonin transporter gene ( 5-HTT) and environmental adversity on the onset of depression in humans has been found in fifteen independent studies. It is supported by evidence from animal experiments, pharmacological challenge and neuroimaging investigations. However, negative findings have been reported in two large samples. We explore reasons for the inconsistencies and suggest means to their resolution. Sample age and gender composition emerge as important factors. While the G × E has been consistently detected in young adult samples, there are contradictory findings in adolescent boys and elderly people. The method of assessment of environmental adversity is also important with detailed interview-based approaches being associated with positive G × E findings. Unresolved issues in the definition of the genotype include the dominance of alleles and influence of other polymorphisms, both in 5-HTT and other genes. Assessment of multiple adverse outcomes, including depression, substance use and self-destructive behaviour is needed to clarify the generalisability of the G × E pathogenic mechanisms. Biological and behavioural intermediate phenotypes are yet to be exploited to understand the mechanisms underlying the G × E.

Symptom dimensions have not yet been comprehensively tested as predictors of the substantial heterogeneity in outcomes of antidepressant treatment in major depressive disorder. We tested nine symptom dimensions derived from a previously published factor analysis of depression rating scales as predictors of outcome in 811 adults with moderate to severe depression treated with flexibly dosed escitalopram or nortriptyline in Genome-based Therapeutic Drugs for Depression (GENDEP). The effects of symptom dimensions were tested in mixed-effect regression models that controlled for overall initial depression severity, age, sex and recruitment centre. Significant results were tested for replicability in 3637 adult out-patients with non-psychotic major depression treated with citalopram in level I of Sequenced Treatment Alternatives to Relieve Depression (STAR*D). The interest-activity symptom dimension (reflecting low interest, reduced activity, indecisiveness and lack of enjoyment) at baseline strongly predicted poor treatment outcome in GENDEP, irrespective of overall depression severity, antidepressant type and outcome measure used. The prediction of poor treatment outcome by the interest-activity dimension was robustly replicated in STAR*D, independent of a comprehensive list of baseline covariates. Loss of interest, diminished activity and inability to make decisions predict poor outcome of antidepressant treatment even after adjustment for overall depression severity and other clinical covariates. The prominence of such symptoms may require additional treatment strategies and should be accounted for in future investigations of antidepressant response.

Molecular genetic analysis offers opportunities to advance our understanding of the nosological relationship between psychiatric diagnostic categories in general, and the mood and psychotic disorders in particular. Strong evidence ( P =7.0 × 10 −7 ) of association at the polymorphism rs1006737 (within CACNA1C, the gene encoding the α-1C subunit of the L-type voltage-gated calcium channel) with the risk of bipolar disorder (BD) has recently been reported in a meta-analysis of three genome-wide association studies of BD, including our BD sample ( N =1868) studied within the Wellcome Trust Case Control Consortium. Here, we have used our UK case samples of recurrent major depression ( N =1196) and schizophrenia ( N =479) and UK non-psychiatric comparison groups ( N =15316) to examine the spectrum of phenotypic effect of the bipolar risk allele at rs1006737. We found that the risk allele conferred increased risk for schizophrenia ( P =0.034) and recurrent major depression ( P =0.013) with similar effect sizes to those previously observed in BD (allelic odds ratio ∼1.15). Our findings are evidence of some degree of overlap in the biological underpinnings of susceptibility to mental illness across the clinical spectrum of mood and psychotic disorders, and show that at least some loci can have a relatively general effect on susceptibility to diagnostic categories, as currently defined. Our findings will contribute to a better understanding of the pathogenesis of major psychiatric illness, and such knowledge should be useful in providing an etiological rationale for shaping psychiatric nosology, which is currently reliant entirely on descriptive clinical data.

Major depressive disorder (MDD) is a common and disabling condition with well-established heritability and environmental risk factors. Gene-environment interaction studies in MDD have typically investigated candidate genes, though the disorder is known to be highly polygenic. This study aims to test for interaction between polygenic risk and stressful life events (SLEs) or childhood trauma (CT) in the aetiology of MDD. The RADIANT UK sample consists of 1605 MDD cases and 1064 controls with SLE data, and a subset of 240 cases and 272 controls with CT data. Polygenic risk scores (PRS) were constructed using results from a mega-analysis on MDD by the Psychiatric Genomics Consortium. PRS and environmental factors were tested for association with case/control status and for interaction between them. PRS significantly predicted depression, explaining 1.1% of variance in phenotype (p = 1.9 × 10(-6)). SLEs and CT were also associated with MDD status (p = 2.19 × 10(-4) and p = 5.12 × 10(-20), respectively). No interactions were found between PRS and SLEs. Significant PRSxCT interactions were found (p = 0.002), but showed an inverse association with MDD status, as cases who experienced more severe CT tended to have a lower PRS than other cases or controls. This relationship between PRS and CT was not observed in independent replication samples. CT is a strong risk factor for MDD but may have greater effect in individuals with lower genetic liability for the disorder. Including environmental risk along with genetics is important in studying the aetiology of MDD and PRS provide a useful approach to investigating gene-environment interactions in complex traits.

It has been well established that both genes and non-shared environment contribute substantially to the underlying aetiology of major depressive disorder (MDD). A comprehensive overview of genetic research in MDD is presented. Method Papers were retrieved from PubMed up to December 2011, using many keywords including: depression, major depressive disorder, genetics, rare variants, gene-environment, whole genome, epigenetics, and specific candidate genes and variants. These were combined in a variety of permutations. Linkage studies have yielded some promising chromosomal regions in MDD. However, there is a continued lack of consistency in association studies, in both candidate gene and genome-wide association studies (GWAS). Numerous factors may account for variable results including the use of different diagnostic approaches, small samples in early studies, population stratification, epigenetic phenomena, copy number variation (CNV), rare variation, and phenotypic and allelic heterogeneity. The conflicting results are also probably, in part, a consequence of environmental factors not being considered or controlled for. Each research group has to identify what issues their sample may best address. We suggest that, where possible, more emphasis should be placed on the environment in molecular behavioural genetics to identify individuals at environmental high risk in addition to genetic high risk. Sequencing should be used to identify rare and alternative variation that may act as a risk factor, and a systems biology approach including gene-gene interactions and pathway analyses would be advantageous. GWAS may require even larger samples with reliably defined (sub)phenotypes.

Response and remission defined by cut-off values on the last observed depression severity score are commonly used as outcome criteria in clinical trials, but ignore the time course of symptomatic change and may lead to inefficient analyses. We explore alternative categorization of outcome by naturally occurring trajectories of symptom change. Growth mixture models were applied to repeated measurements of depression severity in 807 participants with major depression treated for 12 weeks with escitalopram or nortriptyline in the part-randomized Genome-based Therapeutic Drugs for Depression study. Latent trajectory classes were validated as outcomes in drug efficacy comparison and pharmacogenetic analyses. The final two-piece growth mixture model categorized participants into a majority (75%) following a gradual improvement trajectory and the remainder following a trajectory with rapid initial improvement. The rapid improvement trajectory was over-represented among nortriptyline-treated participants and showed an antidepressant-specific pattern of pharmacogenetic associations. In contrast, conventional response and remission favoured escitalopram and produced chance results in pharmacogenetic analyses. Controlling for drop-out reduced drug differences on response and remission but did not affect latent trajectory results. Latent trajectory mixture models capture heterogeneity in the development of clinical response after the initiation of antidepressants and provide an outcome that is distinct from traditional endpoint measures. It differentiates between antidepressants with different modes of action and is robust against bias due to differential discontinuation.

Childhood adversity has been associated with onset of psychosis in adulthood but these studies have used only general definitions of this environmental risk indicator. Therefore, we sought to explore the prevalence of more specific adverse childhood experiences amongst those with and without psychotic disorders using detailed assessments in a large epidemiological case-control sample (AESOP). Data were collected on 182 first-presentation psychosis cases and 246 geographically matched controls in two UK centres. Information relating to the timing and frequency of exposure to different types of childhood adversity (neglect, antipathy, physical and sexual abuse, local authority care, disrupted living arrangements and lack of supportive figure) was obtained using the Childhood Experience of Care and Abuse Questionnaire. Psychosis cases were three times more likely to report severe physical abuse from the mother that commenced prior to 12 years of age, even after adjustment for other significant forms of adversity and demographic confounders. A non-significant trend was also evident for greater prevalence of reported severe maternal antipathy amongst those with psychosis. Associations with maternal neglect and childhood sexual abuse disappeared after adjusting for maternal physical abuse and antipathy. Paternal maltreatment and other forms of adversity were not associated with psychosis nor was there evidence of a dose-response effect. These findings suggest that only specific adverse childhood experiences are associated with psychotic disorders and only in a minority of cases. If replicated, this greater precision will ensure that research into the mechanisms underlying the pathway from childhood adversity to psychosis is more fruitful.

Transcriptional differences in interleukin-11 ( IL11 ) after antidepressant treatment have been found to correspond to clinical response in major depressive disorder (MDD) patients. Expression differences were partly mediated by a single-nucleotide polymorphism (rs1126757), identified as a predictor of antidepressant response as part of a genome-wide association study. Here we attempt to identify whether DNA methylation, another baseline factor known to affect transcription factor binding, might also predict antidepressant response, using samples collected from the Genome-based Therapeutic Drugs for Depression project (GENDEP). DNA samples from 113 MDD individuals from the GENDEP project, who were treated with either escitalopram ( n =80) or nortriptyline ( n =33) for 12 weeks, were randomly selected. Percentage change in Montgomery–Åsberg Depression Rating Scale scores between baseline and week 12 were utilized as our measure of antidepressant response. The Sequenom EpiTYPER platform was used to assess DNA methylation across the only CpG island located in the IL11 gene. Regression analyses were then used to explore the relationship between CpG unit methylation and antidepressant response. We identified a CpG unit predictor of general antidepressant response, a drug by CpG unit interaction predictor of response, and a CpG unit by rs1126757 interaction predictor of antidepressant response. The current study is the first to investigate the potential utility of pharmaco-epigenetic biomarkers for the prediction of antidepressant response. Our results suggest that DNA methylation in IL11 might be useful in identifying those patients likely to respond to antidepressants, and if so, the best drug suited to each individual.