Fronto-limbic structural brain abnormalities have been reported in patients with bipolar disorder (BD), but findings in individuals at increased genetic risk of developing BD have been inconsistent. We conducted a study in adolescents and young adults (12–30 years) comparing measures of fronto-limbic cortical and subcortical brain structure between individuals at increased familial risk of BD (at risk; AR), subjects with BD and controls (CON). We separately examined cortical volume, thickness and surface area as these have distinct neurodevelopmental origins and thus may reflect differential effects of genetic risk.

We compared fronto-limbic measures of grey and white matter volume, cortical thickness and surface area in 72 unaffected-risk individuals with at least one first-degree relative with bipolar disorder (AR), 38 BD subjects and 72 participants with no family history of mental illness (CON).

The AR group had significantly reduced cortical thickness in the left pars orbitalis of the inferior frontal gyrus (IFG) compared with the CON group, and significantly increased left parahippocampal gyral volume compared with those with BD.

The finding of reduced cortical thickness of the left pars orbitalis in AR subjects is consistent with other evidence supporting the IFG as a key region associated with genetic liability for BD. The greater volume of the left parahippocampal gyrus in those at high risk is in line with some prior reports of regional increases in grey matter volume in at-risk subjects. Assessing multiple complementary morphometric measures may assist in the better understanding of abnormal developmental processes in BD.

Using a large Canadian population-based sample, this study aimed to verify whether televiewing in toddlerhood is prospectively associated with self-reported social impairment in middle school.

Participants are from a prospective–longitudinal birth cohort of 991 girls and 1006 boys from the Quebec Longitudinal Study of Child Development. Child self-reported ratings of relational difficulties at age 13 years were linearly regressed on parent-reported televiewing at age 2 years while adjusting for potential confounders.

Every additional 1 h of early childhood television exposure corresponded to an 11% s.d. unit increase in self-reported peer victimization [unstandardized β = 0.03, 95% confidence interval (CI) 0.02–0.04], a 10% s.d. unit increase in self-reported social isolation (unstandardized β = 0.04, 95% CI 0.03–0.05), a 9% s.d. unit increase in self-reported proactive aggression (unstandardized β = 0.02, 95% CI 0.01–0.03) and a 6% s.d. unit increase in self-reported antisocial behavior (unstandardized β = 0.01, 95% CI 0.01–0.01) at age 13 years. These results are above and beyond pre-existing individual and family factors.

Televiewing in toddlerhood was prospectively associated with experiencing victimization and social withdrawal from fellow students and engaging in antisocial behavior and proactive aggression toward fellow students at age 13 years. Adolescents who experience relational difficulties are at risk of long-term health problems (like depression and cardiometabolic disease) and socio-economic problems (like underachievement and unemployment). These relationships, observed more than a decade later, and independent of key potential confounders, suggest a need for better parental awareness of how young children invest their limited waking hours.

There is limited information on long-term outcomes of adolescent depression. This study examines the associations between severity of depression in adolescence and a broad array of adult functional outcomes.

Data were gathered as part of the Christchurch Health and Development Study, a 35-year longitudinal study of a birth cohort of 1265 children born in Christchurch, New Zealand in 1977. Severity of depression at age 14–16 years was classified into three levels according to DSM symptom criteria for major depression (no depression/sub-threshold symptoms/major depression). This classification was related to adult functional outcomes assessed at ages 30 and 35 years using a generalized estimating equation modeling approach. Outcome measures spanned domains of mental disorder, education/economic circumstances, family circumstances and partner relationships.

There were modest but statistically significant bivariate associations between adolescent depression severity and most outcomes. After covariate adjustment there remained weak but significant (p < 0.05) associations with rates of major depression, anxiety disorder, illicit substance abuse/dependence, any mental health problem and intimate partner violence (IPV) victimization. Estimates of attributable risk for these outcomes ranged from 3.8% to 7.8%. For two outcomes there were significant (p < 0.006) gender interactions such that depression severity was significantly related to increased rates of unplanned pregnancy and IPV victimization for females but not for males.

The findings reinforce the importance of the individual/family context in which adolescent depression occurs. When contextual factors and probable maturational effects are taken into account the direct effects of adolescent depression on functioning in mature adulthood appear to be very modest.

Impairments in key neuropsychological domains (e.g. working memory, attention) and social cognitive deficits have been implicated as intermediate (endo) phenotypes for bipolar disorder (BD), and should therefore be evident in unaffected relatives.

Neurocognitive and social cognitive ability was examined in 99 young people (age range 16–30 years) with a biological parent or sibling diagnosed with the disorder [thus deemed to be at risk (AR) of developing BD], compared with 78 healthy control (HC) subjects, and 52 people with a confirmed diagnosis of BD.

Only verbal intelligence and affective response inhibition were significantly impaired in AR relative to HC participants; the BD participants showed significant deficits in attention tasks compared with HCs. Neither AR nor BD patients showed impairments in general intellectual ability, working memory, visuospatial or language ability, relative to HC participants. Analysis of BD-I and BD-II cases separately revealed deficits in attention and immediate memory in BD-I patients (only), relative to HCs. Only the BD (but not AR) participants showed impaired emotion recognition, relative to HCs.

Selective cognitive deficits in the capacity to inhibit negative affective information, and general verbal ability may be intermediate markers of risk for BD; however, the extent and severity of impairment in this sample was less pronounced than has been reported in previous studies of older family members and BD cases. These findings highlight distinctions in the cognitive profiles of AR and BD participants, and provide limited support for progressive cognitive decline in association with illness development in BD.

Delusion is the most characteristic symptom of psychosis. While researchers suggested an association between changes of the cortical salience network (CSN) and delusion, whether these CSN findings are a cause or a consequence of delusion remains unknown.

To assess the effect of CSN functioning to forthcoming changes in delusion scores, we measured brain activation with 3-T functional magnetic resonance imaging in two independent samples of first-episode psychosis patients (total of 27 patients and 23 healthy controls). During scanning, the patients evaluated statements about whether an individual's psychosis-related experiences should be described as a mental illness, and control statements that were also evaluated by healthy controls. Symptoms were assessed at the baseline and at 2 months follow-up with Brief Psychiatric Rating Scale.

Both tasks activated the CSN in comparison with rest. Activation of CSN (‘illness evaluation v. control task’ contrast) in patients positively correlated with worsening of or less improvement in delusions at the 2-month follow-up assessment. This finding was independent of delusion and clinical insight scores at the baseline evaluation.

Our findings link symptom-evaluation-related CSN functioning to severity of delusion and, importantly, add a new layer of evidence for the contribution of CSN functioning to the longitudinal course of delusions.

Patients with a deletion at chromosome 22q11.2 (22q11DS) have 30% lifetime risk of developing a psychosis. People fulfilling clinical criteria for ultra-high risk (UHR) for psychosis have 30% risk of developing a psychosis within 2 years. Both high-risk groups show white-matter (WM) abnormalities in microstructure and volume compared to healthy controls (HC), which have been related to psychotic symptoms. Comparisons of WM pathology between these two groups may specify WM markers related to genetic and clinical risk factors.

Fractional anisotropy (FA), axial diffusivity (AD), radial diffusivity (RD) and mean diffusivity (MD) were assessed using diffusion tensor magnetic resonance imaging (MRI), and WM volume with structural MRI, in 23 UHR patients, 21 22q11DS patients, and 33 HC.

Compared to UHR patients 22q11DS patients had (1) lower AD and RD in corpus callosum (CC), cortical fasciculi, and anterior thalamic radiation (ATR), (2) higher FA in CC and ATR, and (3) lower occipital and superior temporal gyrus WM volume. Compared to HC, 22q11DS patients had (1) lower AD and RD throughout cortical fasciculi and (2) higher FA in ATR, CC and inferior fronto-occipital fasciculus. Compared to HC, UHR patients had (1) higher mean MD, RD, and AD in CC, ATR and cortical fasciculi, (2) no differences in FA.

UHR and 22q11DS patients share a susceptibility for developing psychosis yet were characterized by distinct patterns of WM alterations relative to HC. While UHR patients were typified by signs suggestive of aberrant myelination, 22q11DS subjects showed signs suggestive of lower axonal integrity.

Twin studies have been criticized for upwardly biased estimates that might contribute to the missing heritability problem.

We identified, from the general Swedish population born 1960–1990, informative sibships containing a proband, one reared-together full- or half-sibling and a full-, step- or half-sibling with varying degrees of childhood cohabitation with the proband. Estimates of genetic, shared and individual specific environment for drug abuse (DA), alcohol use disorder (AUD) and criminal behavior (CB), assessed from medical, legal or pharmacy registries, were obtained using Mplus.

Aggregate estimates of additive genetic effects for DA, AUD and CB obtained separately in males and females varied from 0.46 to 0.73 and agreed with those obtained from monozygotic and dizygotic twins from the same population. Of 54 heritability estimates from individual classes of informative sibling trios (3 syndromes × 9 classes of trios × 2 sexes), heritability estimates from the siblings were lower, tied and higher than those from obtained from twins in 26, one and 27 comparisons, respectively. By contrast, of 54 shared environmental estimates, 33 were lower than those found in twins, one tied and 20 were higher.

With adequate information, human populations can provide many methods for estimating genetic and shared environmental effects. For the three externalizing syndromes examined, concerns that heritability estimates from twin studies are upwardly biased or were not generalizable to more typical kinds of siblings were not supported. Overestimation of heritability from twin studies is not a likely explanation for the missing heritability problem.

Social anxiety disorder involves fear of social objects or situations. Social referencing may play an important role in the acquisition of this fear and could be a key determinant in future biomarkers and treatment pathways. However, the neural underpinnings mediating such learning in social anxiety are unknown. Using event-related functional magnetic resonance imaging, we examined social reference learning in social anxiety disorder. Specifically, would patients with the disorder show increased amygdala activity during social reference learning, and further, following social reference learning, show particularly increased response to objects associated with other people's negative reactions?

A total of 32 unmedicated patients with social anxiety disorder and 22 age-, intelligence quotient- and gender-matched healthy individuals responded to objects that had become associated with others’ fearful, angry, happy or neutral reactions.

During the social reference learning phase, a significant group × social context interaction revealed that, relative to the comparison group, the social anxiety group showed a significantly greater response in the amygdala, as well as rostral, dorsomedial and lateral frontal and parietal cortices during the social, relative to non-social, referencing trials. In addition, during the object test phase, relative to the comparison group, the social anxiety group showed increased bilateral amygdala activation to objects associated with others’ fearful reactions, and a trend towards decreased amygdala activation to objects associated with others’ happy and neutral reactions.

These results suggest perturbed observational learning in social anxiety disorder. In addition, they further implicate the amygdala and dorsomedial prefrontal cortex in the disorder, and underscore their importance in future biomarker developments.

Functional remediation is a novel intervention with demonstrated efficacy at improving functional outcome in euthymic bipolar patients. However, in a previous trial no significant changes in neurocognitive measures were detected. The objective of the present analysis was to test the efficacy of this therapy in the enhancement of neuropsychological functions in a subgroup of neurocognitively impaired bipolar patients.

A total of 188 out of 239 DSM-IV euthymic bipolar patients performing below two standard deviations from the mean of normative data in any neurocognitive test were included in this subanalysis. Repeated-measures analyses of variance were conducted to assess the impact of the treatment arms [functional remediation, psychoeducation, or treatment as usual (TAU)] on participants’ neurocognitive and functional outcomes in the subgroup of neurocognitively impaired patients.

Patients receiving functional remediation (n = 56) showed an improvement on delayed free recall when compared with the TAU (n = 63) and psychoeducation (n = 69) groups as shown by the group × time interaction at 6-month follow-up [F2,158 = 3.37, degrees of freedom (df) = 2, p = 0.037]. However, Tukey post-hoc analyses revealed that functional remediation was only superior when compared with TAU (p = 0.04), but not with psychoeducation (p = 0.10). Finally, the patients in the functional remediation group also benefited from the treatment in terms of functional outcome (F2,158 = 4.26, df = 2, p = 0.016).

Functional remediation is effective at improving verbal memory and psychosocial functioning in a sample of neurocognitively impaired bipolar patients at 6-month follow-up. Neurocognitive enhancement may be one of the active ingredients of this novel intervention, and, specifically, verbal memory appears to be the most sensitive function that improves with functional remediation.

Certain personality traits predispose to criminal behavior (CB). We further clarify this relationship in a Swedish national sample.

Psychological strength (PS) was assessed on a nine-point scale at personal interview in 1 653 721 Swedish men aged 18–20 years. We examined the association between PS and total, violent and recurrent CB over the lifetime (logistic regression), prospectively (Cox regression) and by bivariate Cholesky decomposition in 2507 monozygotic and 2244 dizygotic twin pairs (OpenMx).

Examining linear effects by logistic regression, PS was robustly associated with lifetime risk of total CB (per point, odds ratio = 0.74) and even more strongly associated with risk for violent (0.69) and recurrent CB (0.52). Prospective predictions of these three forms of CB by PS were similar, with hazard ratios of 0.80, 0.73 and 0.54, respectively. Twin modeling demonstrated that, for all three CB types, the association with PS arose almost entirely from familial effects. Common shared environment accounted for 72, 56 and 43% of the phenotypic correlation between PS and, respectively, total, violent and recurrent CB. Parallel figures for common genetic effects were for 24, 37 and 54%, respectively.

PS is strongly related to risk for total CB, and even more strongly for violent and, especially, recurrent CB. This association is probably not causal but rather results from shared familial risk factors that make an impact both on PS and risk for CB. PS has a stronger overall correlation with more severe criminal outcomes and a higher proportion of that correlation results from common genetic factors.

Risk factors for alcohol problems (AP) include biological and environmental factors that are relevant across development. The pathways through which these factors are related, and how they lead to AP, are optimally considered in the context of a comprehensive developmental model.

Using data from a prospectively assessed, population-based UK cohort, we constructed a structural equation model that integrated risk factors reflecting individual, family and peer/community-level constructs across childhood, adolescence and young adulthood. These variables were used to predict AP at the age of 20 years.

The final model explained over 30% of the variance in liability to age 20 years AP. Most prominent in the model was an externalizing pathway to AP, with conduct problems, sensation seeking, AP at age 17.5 years and illicit substance use acting as robust predictors. In conjunction with these individual-level risk factors, familial AP, peer relationships and low parental monitoring also predicted AP. Internalizing problems were less consistently associated with AP. Some risk factors previously identified were not associated with AP in the context of this comprehensive model.

The etiology of young adult AP is complex, influenced by risk factors that manifest across development. The most prominent pathway to AP is via externalizing and related behaviors. These findings underscore the importance of jointly assessing both biologically influenced and environmental risk factors for AP in a developmental context.

Whether the incidence of eating disorders in Western, industrialized countries has changed over time has been the subject of much debate. The purpose of this primary-care study was to examine changes in the incidence of eating disorders in The Netherlands during the 1980s, 1990s and 2000s.

A nationwide network of general practitioners (GPs), serving a representative sample (~1%) of the total Dutch population, recorded newly diagnosed patients with anorexia nervosa (AN) and bulimia nervosa (BN) in their practice during 1985–1989, 1995–1999, and 2005–2009. GPs are key players in the Dutch healthcare system, as their written referral is mandatory in order to get access to specialized (mental) healthcare, covered by health insurance. Health insurance is virtually universal in The Netherlands (99% of the population). A substantial number of GPs participated in all three study periods, during which the same case identification criteria were used and the same psychiatrist was responsible for making the final diagnoses. Incidence rates were calculated and for comparison between periods, incidence rate ratios.

The overall incidence rate of BN decreased significantly in the past three decades (from 8.6 per 100 000 person-years in 1985–1989 to 6.1 in 1995–1999, and 3.2 in 2005–2009). The overall incidence of AN remained fairly stable during three decades, i.e. 7.4 per 1 00 000 person-years in 1985–1989, 7.8 in 1995–1999, and 6.0 in 2005–2009.

The incidence rate of BN decreased significantly over the past three decades, while the overall incidence rate of AN remained stable.

Cannabis use disorder is associated with an earlier age at onset and a more severe outcome of schizophrenia spectrum disorders. The role of cannabis use before the onset of illness (premorbid cannabis use) has not been fully investigated. We here examined how amount and type of premorbid cannabis use was associated with the later course of illness including current substance use, symptoms and level of functioning in schizophrenia spectrum disorder.

We used a naturalistic sample of patients with DSM-IV schizophrenia spectrum disorders with a comprehensive history of illness and substance use. Data on premorbid substance use was obtained from comprehensive self-report. The relationship to outcome was investigated using regression models that included current substance use and premorbid functioning.

Pre-schizophrenia cannabis use was significantly associated with more severe psychotic symptoms and impaired functioning. Higher levels of premorbid cannabis use were associated with higher levels of current psychotic symptoms. These associations were independent of current substance use and premorbid functioning. Early use of cannabis (age <17 years) was associated with earlier age at onset of psychosis, independently of potential confounders.

Pre-psychosis cannabis use affects illness outcome in schizophrenia spectrum disorders, and is associated with lower age at onset of psychosis. These findings of independent negative effects of premorbid cannabis use in schizophrenia suggest that a limitation of the general use of cannabis may have beneficial health effects.

Intrusive memories of traumatic events are a core feature of post-traumatic stress disorder but little is known about the neurobiological formation of intrusions. The aim of this study was to determine whether the activity of the noradrenergic system during an intrusion-inducing stressor would influence subsequent intrusive memories.

We conducted an experimental, double-blind, placebo-controlled study in 118 healthy women. Participants received a single dose of either 10 mg yohimbine, stimulating noradrenergic activity, or 0.15 mg clonidine, inhibiting noradrenergic activity, or placebo. Subsequently, they watched an established trauma film which induced intrusions. The number of consecutive intrusions resulting from the trauma film, the vividness of the intrusions, and the degree of distress evoked by the intrusions were assessed during the following 4 days. Salivary cortisol and α-amylase were collected before and after the trauma film.

A significant time × treatment interaction for the number of intrusions and the vividness of intrusions indicated a different time course of intrusions depending on treatment. Post-hoc tests revealed a delayed decrease of intrusions and a delayed decrease of intrusion vividness after the trauma film in the yohimbine group compared with the clonidine and placebo groups. Furthermore, after yohimbine administration, a significant increase in salivary cortisol levels was observed during the trauma film.

Our findings indicate that pharmacological activation of the noradrenergic system during an emotionally negative event makes an impact on consecutive intrusive memories and their vividness in healthy women. The noradrenergic system seems to be involved in the formation of intrusive memories.

The nature of the relationship between depressive vulnerability (DV) and acute adversity in the etiology of major depression (MD) remains poorly understood.

Stressful life events (SLEs) and MD onsets in the last year were assessed at four waves in cohort 1 (females) and at two waves in cohort 2 (males and females) from the Virginia Adult Twin Study. Structural equation modeling was conducted in Mplus.

In cohort 1, DV was strongly indexed by depressive episodes over the four waves (paths from +0.72 to 0.79) and predicted by SLEs in the month of their occurrence (+0.31 to 0.36). Wave-specific DV was associated both with stable DV (+0.29 to 0.33) and by forward transmission of DV from the preceding wave (+0.33 to 0.36). SLEs were predicted by stable DV (+0.29) and from SLEs in the preceding month (+0.06). As the cohort aged, MD onsets were better indexed by DV and more poorly predicted by SLEs. Parameter estimates were similar in males and females from cohort 2. In individuals with prior depressive episodes, the association between MD onset and SLEs was weakened while the prediction of SLEs from DV was substantially strengthened. We found no evidence for ‘reverse causation’ from MD episodes to SLEs.

The interrelationship between DV and acute adversity in the etiology of MD is complex and temporally dynamic. DV impacts on MD risk both directly and indirectly through selection into high stress environments. Over time, depressive episodes become more autonomous. Both DV and SLEs transmit forward over time and therefore form clear targets for intervention.

A hallmark symptom after psychological trauma is the presence of intrusive memories. It is unclear why only some moments of trauma become intrusive, and how these memories involuntarily return to mind. Understanding the neural mechanisms involved in the encoding and involuntary recall of intrusive memories may elucidate these questions.

Participants (n = 35) underwent functional magnetic resonance imaging (fMRI) while being exposed to traumatic film footage. After film viewing, participants indicated within the scanner, while undergoing fMRI, if they experienced an intrusive memory of the film. Further intrusive memories in daily life were recorded for 7 days. After 7 days, participants completed a recognition memory test. Intrusive memory encoding was captured by comparing activity at the time of viewing ‘Intrusive scenes’ (scenes recalled involuntarily), ‘Control scenes’ (scenes never recalled involuntarily) and ‘Potential scenes’ (scenes recalled involuntarily by others but not that individual). Signal change associated with intrusive memory involuntary recall was modelled using finite impulse response basis functions.

We found a widespread pattern of increased activation for Intrusive v. both Potential and Control scenes at encoding. The left inferior frontal gyrus and middle temporal gyrus showed increased activity in Intrusive scenes compared with Potential scenes, but not in Intrusive scenes compared with Control scenes. This pattern of activation persisted when taking recognition memory performance into account. Intrusive memory involuntary recall was characterized by activity in frontal regions, notably the left inferior frontal gyrus.

The left inferior frontal gyrus may be implicated in both the encoding and involuntary recall of intrusive memories.

Eating disorders (EDs) have long-term physical and mental impacts on those affected. However, few population-based studies have estimated the prevalence of EDs. We aimed to estimate the lifetime and 12-month prevalence rates of EDs using DSM-IV criteria, and to examine differences against the DSM-5 criteria for anorexia.

A nationally representative sample of 10 038 residents in Switzerland was interviewed, and prevalence rates for anorexia nervosa (AN), bulimia nervosa (BN), and binge eating disorder (BED) were assessed using WHO Composite International Diagnostic Interviews (WHO-CIDI).

The lifetime prevalence rate for any ED was found to be 3.5%. Lifetime prevalence estimates for AN, BN, and/or BED were 1.2%, 2.4%, and 2.4%, respectively, among women and 0.2%, 0.9%, and 0.7%, respectively, among men. Utilizing the DSM-5 criteria, the prevalence of AN in women increased by more than 50%, from 1.2% to 1.9%. Among those meeting the criteria for any ED, only 49.4% of men and 67.9% of women had ever sought professional help about their problems with eating or weight.

The higher prevalence of BN we detected relative to other studies should prompt further monitoring for a possible increasing trend. The female v. male ratios, especially for bulimia and BED, are decreasing. Given that more than half of those affected have never consulted any professional about their problems with eating or weight, routine inquiries about eating and weight by clinicians, school teachers/psychologists, and family members may help those who are at risk, especially among men.

Healthy lifestyles prevent cardiovascular disease and are increasingly recognized in relation to mental health but longitudinal studies are limited. We examined bi-directional associations between mood disorders and healthy lifestyles in a cohort followed for 5 years.

Participants were aged 26–36 years at baseline (2004–2006) and 31–41 years at follow-up (2009–2011). At follow-up, lifetime mood disorders (depression or dysthymia) were retrospectively diagnosed with the Composite International Diagnostic Interview. A five-item lifestyle score (comprising body mass index, non-smoking, alcohol consumption, leisure time physical activity and healthy diet) was measured at both time points. Linear and log multinomial regression determined if mood disorder before baseline predicted changes in lifestyle (n = 1041). Log binomial regression estimated whether lifestyle at baseline predicted new episodes of mood disorder (n = 1233). Covariates included age, sex, socio-economic position, parental and marital status, social support, major life events, cardiovascular disease history, and self-rated physical and mental health.

A history of mood disorder before baseline predicted unfavourable trajectories of lifestyle over follow-up, including somewhat lower risk of improvement [relative risk (RR) 0.76, 95% confidence interval (CI) 0.56–1.03] and greater risk of worsening (RR 1.46, 95% CI 0.99–2.15) of lifestyle independent of confounding factors. Higher lifestyle scores at baseline were associated with a 22% (RR 0.76, 95% CI 0.61–0.95) reduced risk of first episodes of mood disorder, independent of confounding factors.

Healthy lifestyles and mood disorders are closely related. Our results suggest that healthy lifestyles may not only reduce cardiovascular disease but also promote mental health.

Given the frequency with which families change residences, the effects of childhood relocations have gained increasing research attention. Many researchers have demonstrated that childhood relocations are associated with a variety of adverse outcomes. However, drawing strong causal claims remains problematic due to uncontrolled confounding factors.

We utilized longitudinal, population-based Swedish registers to generate a nationally representative sample of offspring born 1983–1997 (n = 1 510 463). Using Cox regression and logistic regression, we examined the risk for numerous adverse outcomes after childhood relocation while controlling for measured covariates. To account for unmeasured genetic and environmental confounds, we also compared differentially exposed cousins and siblings.

In the cohort baseline model, each annual relocation was associated with risk for the adverse outcomes, including suicide attempt [hazard ratio (HR) 1.19, 95% confidence interval (CI) 1.19–1.20]. However, when accounting for offspring and parental covariates (HR 1.08, 95% CI 1.07–1.09), as well as genetic and environmental confounds shared by cousins (HR 1.07, 95% CI 1.05–1.09) and siblings (HR 1.00, 95% CI 0.97–1.04), the risk for suicide attempt attenuated. We found a commensurate pattern of results for severe mental illness, substance abuse, criminal convictions, and low academic achievement.

Previous research may have overemphasized the independent association between relocations and later adverse outcomes. The results suggest that the association between childhood relocations and suicide attempt, psychiatric problems, and low academic achievement is partially explained by genetic and environmental confounds correlated with relocations. This study demonstrates the importance of using family-based, quasi-experimental designs to test plausible alternate hypotheses when examining causality.

Although interventions exist to reduce violent crime, optimal implementation requires accurate targeting. We report the results of an attempt to develop an actuarial model using machine learning methods to predict future violent crimes among US Army soldiers.

A consolidated administrative database for all 975 057 soldiers in the US Army in 2004–2009 was created in the Army Study to Assess Risk and Resilience in Servicemembers (Army STARRS). Of these soldiers, 5771 committed a first founded major physical violent crime (murder-manslaughter, kidnapping, aggravated arson, aggravated assault, robbery) over that time period. Temporally prior administrative records measuring socio-demographic, Army career, criminal justice, medical/pharmacy, and contextual variables were used to build an actuarial model for these crimes separately among men and women using machine learning methods (cross-validated stepwise regression, random forests, penalized regressions). The model was then validated in an independent 2011–2013 sample.

Key predictors were indicators of disadvantaged social/socioeconomic status, early career stage, prior crime, and mental disorder treatment. Area under the receiver-operating characteristic curve was 0.80–0.82 in 2004–2009 and 0.77 in the 2011–2013 validation sample. Of all administratively recorded crimes, 36.2–33.1% (male-female) were committed by the 5% of soldiers having the highest predicted risk in 2004–2009 and an even higher proportion (50.5%) in the 2011–2013 validation sample.

Although these results suggest that the models could be used to target soldiers at high risk of violent crime perpetration for preventive interventions, final implementation decisions would require further validation and weighing of predicted effectiveness against intervention costs and competing risks.