Depression is common in chronic illness and screening for depression has been widely recommended. There have been no large studies of screening for depression in routine care for patients with chronic illness.

We performed a retrospective cohort study to examine the timing of new depression diagnosis or treatment in relation to annual screening for depression in patients with coronary heart disease (CHD) or diabetes. We examined a database derived from 1.3 million patients registered with general practices in Scotland for the year commencing 1 April 2007. Eligible patients had either CHD or diabetes, were screened for depression during the year and either received a new diagnosis of depression or commenced a new course of antidepressant (excluding those commonly used to treat diabetic neuropathy). Analysis was by the self-controlled case-series method with the outcome measure being the relative incidence (RI) in the period 1–28 days after screening compared to other times.

A total of 67358 patients were screened for depression and 2269 received a new diagnosis or commenced treatment. For the period after screening, the RI was 3.03 [95% confidence interval (CI) 2.44–3.78] for diagnosis and 1.78 (95% CI 1.54–2.05) for treatment. The number needed to screen was 976 (95% CI 886–1104) for a new diagnosis and 687 (95% CI 586–853) for new antidepressant treatment.

Systematic screening for depression in patients with chronic disease in primary care results in a significant but small increase in new diagnosis and treatment in the following 4 weeks.

Although patients with chronic schizophrenia have substantially higher smoking rates than either the general population or patients with other mental illnesses, drug-naive patients with a first episode of schizophrenia have received little systemic study. This study examined smoking rates, the association between smoking and symptom severity and cognitive function in Chinese first-episode schizophrenia (FES) patients using cross-sectional and case-control designs.

Two hundred and forty-four drug-naive FES patients and 256 healthy controls matched for gender, age and education completed the Fagerström Test for Nicotine Dependence (FTND) and the Repeatable Battery for the Assessment of Neuropsychological Status (RBANS). Patients were also rated on the Positive and Negative Symptom Scale (PANSS).

The rate and quantity of smoking were not significantly higher among FES patients compared to the general population. Among patients, smokers scored higher than non-smokers on the total PANSS and the positive symptom subscale scores. There were no significant associations between cognitive function and smoking in either FES patients or healthy controls.

In contrast to studies in patients with chronic schizophrenia, drug-naive FES patients did not smoke more frequently than the general population. Furthermore, patients with psychotic disorders who smoked did not exhibit significant cognitive differences compared with those who did not smoke. However, smoking may have other detrimental effects on physical and mental health, for example on positive symptoms.

Childhood psychotic symptoms have been used as a subclinical phenotype of schizophrenia in etiological research and as a target for preventative interventions. However, recent studies have cast doubt on the specificity of these symptoms for schizophrenia, suggesting alternative outcomes such as anxiety and depression. Using a prospective longitudinal birth cohort we investigated whether childhood psychotic symptoms predicted a diagnosis of schizophrenia or other psychiatric disorders by 38 years of age.

Participants were drawn from a birth cohort of 1037 children from Dunedin, New Zealand, who were followed prospectively to 38 years of age (96% retention rate). Structured clinical interviews were administered at age 11 to assess psychotic symptoms and study members underwent psychiatric assessments at ages 18, 21, 26, 32 and 38 to obtain past-year DSM-III-R/IV diagnoses and self-reports of attempted suicides since adolescence.

Psychotic symptoms at age 11 predicted elevated rates of research diagnoses of schizophrenia and post-traumatic stress disorder (PTSD) and also suicide attempts by age 38, even when controlling for gender, social class and childhood psychopathology. No significant associations were found for persistent anxiety, persistent depression, mania or persistent substance dependence. Very few of the children presenting with age-11 psychotic symptoms were free from disorder by age 38.

Childhood psychotic symptoms were not specific to a diagnosis of schizophrenia in adulthood and thus future studies of early symptoms should be cautious in extrapolating findings only to this clinical disorder. However, these symptoms may be useful as a marker of adult mental health problems more broadly.

Depression is a clinically relevant dimension, associated with both positive and negative symptoms, in patients with schizophrenia. However, in siblings it is unknown whether depression is associated with subclinical positive and negative symptoms.

Depressive symptoms and their association with positive and negative symptoms were examined in 813 healthy siblings of patients with a non-affective psychotic disorder, 822 patients and 527 healthy controls. Depressive episodes meeting DSM-IV-TR criteria (lifetime) and depressed mood (lifetime) were assessed with the Comprehensive Assessment of Symptoms and History (CASH) in all three groups. In the patient group, the severity of positive and negative psychosis symptoms was assessed with the CASH. In the siblings and healthy controls, the severity of subclinical psychosis symptoms was assessed with the Community Assessment of Psychic Experiences (CAPE).

Patients reported more lifetime depressed mood and more depressive episodes than both siblings and controls. Siblings had a higher chance of meeting lifetime depressive episodes than the controls; no significant differences in depressed mood were found between siblings and controls. In all three groups the number and duration of depressive symptoms were associated with (sub)clinical negative symptoms. In the patients and siblings the number of depressive symptoms was furthermore associated with (sub)clinical positive symptoms. Finally, lifetime depressed mood showed familial clustering but this clustering was absent for lifetime depressive episodes.

These findings suggest that a co-occurring genetic vulnerability for both depressive and psychotic symptomatology exists on a clinical and a subclinical level.

Abnormalities in hippocampal–parahippocampal (H-PH) function are prominent features of schizophrenia and have been associated with deficits in episodic memory. However, it remains unclear whether these abnormalities represent a phenotype related to genetic risk for schizophrenia or whether they are related to disease state.

We investigated H-PH-mediated behavior and physiology, using blood oxygenation level-dependent functional magnetic resonance imaging (BOLD fMRI), during episodic memory in a sample of patients with schizophrenia, clinically unaffected siblings and healthy subjects.

Patients with schizophrenia and unaffected siblings displayed abnormalities in episodic memory performance. During an fMRI memory encoding task, both patients and siblings demonstrated a similar pattern of reduced H-PH engagement compared with healthy subjects.

Our findings suggest that the pathophysiological mechanism underlying the inability of patients with schizophrenia to properly engage the H-PH during episodic memory is related to genetic risk for the disorder. Therefore, H-PH dysfunction can be assumed as a schizophrenia susceptibility-related phenotype.

Large epidemiological studies are needed to better understand the prevalence and profile of offending by people with mental illness. This study used a whole-of-population design to examine the prevalence, type and pattern of offending across all psychiatric diagnoses, including schizophrenia, compared to the general population.

We used whole-of-population longitudinal record-linked data for a cohort of all Western Australians born 1955–1969 to determine arrest history over the period 1985–1996 and to ascertain recorded history of psychiatric illness. Of the cohort, 116 656 had been arrested and 40 478 were on the psychiatric case register.

The period prevalence of arrest for people with any psychiatric illness was 32.1%. The highest arrest prevalence, by diagnostic category, was for substance use disorders (59.4%); the prevalence for schizophrenia was 38.7%. Co-morbid substance use disorders significantly increased risk of arrest in people with schizophrenia. The prevalence of mental illness among offenders was 11.1%: 6.5% of offenders had substance use disorders and 1.7% had schizophrenia. For the majority of offenders with a psychiatric illness, first arrest preceded first contact with mental health services; for schizophrenia only, this proportion was increasing over time. The mean percentage annual change in the number of arrests during 1985–1996 rose significantly for offenders with a psychiatric illness other than schizophrenia and dropped significantly for those with no mental illness. Compared to non-psychiatric offenders, offenders with schizophrenia were more likely to offend alone, to offend in open places and to target strangers.

Our findings open the way to an informed approach to the management of offenders with mental illness.

Clinical depression is associated with reductions in white-matter integrity in several long tracts of the brain. The extent to which these findings are localized or related to depressive symptoms or personality traits linked to disease risk remains unclear.

Members of the Lothian Birth Cohort 1936 (LBC936) were assessed in two waves at mean ages of 70 and 73 years. At wave 1, they underwent assessments of depressive symptoms and the personality traits of neuroticism and extraversion. Brain diffusion magnetic resonance imaging (MRI) data were obtained at the second wave and mood assessments were repeated. We tested whether depressive symptoms were related to reduced white-matter tract fractional anisotropy (FA), a measure of integrity, and then examined whether high neuroticism or low extraversion mediated this relationship.

Six hundred and sixty-eight participants provided useable data. Bilateral uncinate fasciculus FA was significantly negatively associated with depressive symptoms at both waves (standardized β=0.12–0.16). Higher neuroticism and lower extraversion were also significantly associated with lower uncinate FA bilaterally (standardized β=0.09–0.15) and significantly mediated the relationship between FA and depressive symptoms.

Trait liability to depression and depressive symptoms are associated with reduced structural connectivity in tracts connecting the prefrontal cortex with the amygdala and anterior temporal cortex. These effects suggest that frontotemporal disconnection is linked to the etiology of depression, in part through personality trait differences.

This study aimed to determine whether depression in patients with long-term conditions is associated with the number of morbidities or the type of co-morbidity.

A cohort study of 299 912 participants aged 30–100 years. The prevalence of depression, rates of health-care utilization and costs were evaluated in relation to diagnoses of diabetes mellitus (DM), coronary heart disease (CHD), stroke and colorectal cancer.

The age-standardized prevalence of depression was 7% in men and 14% in women with no morbidity. The frequency of depression increased in single morbidities including DM (men 13%, women 22%), CHD (men 15%, women 24%), stroke (men 14%, women 26%) or colorectal cancer (men 10%, women 21%). Participants with concurrent diabetes, CHD and stroke had a very high prevalence of depression (men 23%, women 49%). The relative rate of depression for one morbidity was 1.63 [95% confidence interval (CI) 1.59–1.66], two morbidities 1.96 (95% CI 1.89–2.03) and three morbidities 2.35 (95% CI 2.03–2.59). Compared to those with no morbidity, depression was associated with higher rates of health-care utilization and increased costs at any level of morbidity. In women aged 55 to 64 years without morbidity, the mean annual health-care cost was £513 without depression and £1074 with depression; when three morbidities were present, the cost was £1495 without depression and £2878 with depression.

Depression prevalence and health-care costs are more strongly associated with the number of morbidities than the nature of the co-morbid diagnosis.

The longitudinal neuropsychological study of first-episode early-onset psychosis (EOP) patients, whose brain maturation is still in progress at the time of illness onset, provides a unique opportunity to compare their cognitive development with that of healthy subjects, in search of specific patterns resulting from the interaction between neurodevelopmental processes and the presence of psychotic disorders.

Seventy-five first-episode EOP patients (schizophrenia n = 35; bipolar disorder n = 17; other forms of psychosis n = 23) with a mean age of 15.53 years were assessed with a neuropsychological battery that included measures of attention, working memory, memory and executive functions within 6 months following the onset of the first psychotic symptom (baseline) and 2 years later. Psychotic symptoms were assessed at both times with the Positive and Negative Symptom Scale (PANSS). Seventy-nine healthy subjects matched for age and education served as controls.

EOP patients showed significant cognitive impairment at both baseline and the 2-year follow-up, with no significant differences between diagnostic groups at either time. Both healthy controls and EOP patients improved in all cognitive measures, except for patient working memory. Improvement in patient attention lost significance after controlling for psychotic symptom reduction. No significant time/diagnosis interaction was found among patients (p > 0.405).

Cognitive impairment in EOP is already present at the first episode, and cognitive development seems to be arrested early in EOP patients compared to their healthy peers, at least for some cognitive functions. These and previous similar results support the neurodevelopmental hypothesis of psychosis.

Executive dysfunction, distinct from other cognitive deficits in depression, has been associated with suicidal behavior. However, this dysfunction is not found consistently across samples.

Medication-free subjects with DSM-IV major depressive episode (major depressive disorder and bipolar type I disorder) and a past history of suicidal behavior (n = 72) were compared to medication-free depressed subjects with no history of suicidal behavior (n = 80) and healthy volunteers (n = 56) on a battery of tests assessing neuropsychological functions typically affected by depression (motor and psychomotor speed, attention, memory) and executive functions reportedly impaired in suicide attempters (abstract/contingent learning, working memory, language fluency, impulse control).

All of the depressed subjects performed worse than healthy volunteers on motor, psychomotor and language fluency tasks. Past suicide attempters, in turn, performed worse than depressed non-attempters on attention and memory/working memory tasks [a computerized Stroop task, the Buschke Selective Reminding Task (SRT), the Benton Visual Retention Test (VRT) and an N-back task] but not on other executive function measures, including a task associated with ventral prefrontal function (Object Alternation). Deficits were not accounted for by current suicidal ideation or the lethality of past attempts. A small subsample of those using a violent method in their most lethal attempt showed a pattern of poor executive performance.

Deficits in specific components of attention control, memory and working memory were associated with suicidal behavior in a sample where non-violent attempt predominated. Broader executive dysfunction in depression may be associated with specific forms of suicidal behavior, rather than suicidal behavior per se.

This study aimed to determine whether patients with post-traumatic stress disorder (PTSD) show difficulty in recruitment of the regions of the frontal and parietal cortex implicated in top-down attentional control in the presence and absence of emotional distracters.

Unmedicated individuals with PTSD (n = 14), and age-, IQ- and gender-matched individuals exposed to trauma (n = 15) and healthy controls (n = 19) were tested on the affective number Stroop task. In addition, blood oxygen level-dependent responses, as measured via functional magnetic resonance imaging, were recorded.

Patients with PTSD showed disrupted recruitment of lateral regions of the superior and inferior frontal cortex as well as the parietal cortex in the presence of negative distracters. Trauma-comparison individuals showed indications of a heightened ability to recruit fronto-parietal regions implicated in top-down attentional control across distracter conditions.

These results are consistent with suggestions that emotional responsiveness can interfere with the recruitment of regions implicated in top-down attentional control; the heightened emotional responding of patients with PTSD may lead to the heightened interference in the recruitment of these regions.

Schizophrenia is associated with various brain structural abnormalities, including reduced volume of the hippocampi, prefrontal lobes and thalami. Cannabis use increases the risk of schizophrenia but reports of brain structural abnormalities in the cannabis-using population have not been consistent. We used automated image analysis to compare brain structural changes over time in people at elevated risk of schizophrenia for familial reasons who did and did not use cannabis.

Magnetic resonance imaging (MRI) scans were obtained from subjects at high familial risk of schizophrenia at entry to the Edinburgh High Risk Study (EHRS) and approximately 2 years later. Differential grey matter (GM) loss in those exposed (n = 23) and not exposed to cannabis (n = 32) in the intervening period was compared using tensor-based morphometry (TBM).

Cannabis exposure was associated with significantly greater loss of right anterior hippocampal (pcorrected = 0.029, t = 3.88) and left superior frontal lobe GM (pcorrected = 0.026, t = 4.68). The former finding remained significant even after the exclusion of individuals who had used other drugs during the inter-scan interval.

Using an automated analysis of longitudinal data, we demonstrate an association between cannabis use and GM loss in currently well people at familial risk of developing schizophrenia. This observation may be important in understanding the link between cannabis exposure and the subsequent development of schizophrenia.

Group-level results suggest that relative to healthy controls (HCs), ultra-high-risk (UHR) and first-episode psychosis (FEP) subjects show alterations in neuroanatomy, neurofunction and cognition that may be mediated genetically. It is unclear, however, whether these groups can be differentiated at single-subject level, for instance using the machine learning analysis support vector machine (SVM). Here, we used a multimodal approach to examine the ability of structural magnetic resonance imaging (sMRI), functional MRI (fMRI), diffusion tensor neuroimaging (DTI), genetic and cognitive data to differentiate between UHR, FEP and HC subjects at the single-subject level using SVM.

Three age- and gender-matched SVM paired comparison groups were created comprising 19, 19 and 15 subject pairs for FEP versus HC, UHR versus HC and FEP versus UHR, respectively. Genetic, sMRI, DTI, fMRI and cognitive data were obtained for each participant and the ability of each to discriminate subjects at the individual level in conjunction with SVM was tested.

Successful classification accuracies (p < 0.05) comprised FEP versus HC (genotype, 67.86%; DTI, 65.79%; fMRI, 65.79% and 68.42%; cognitive data, 73.69%), UHR versus HC (sMRI, 68.42%; DTI, 65.79%), and FEP versus UHR (sMRI, 76.67%; fMRI, 73.33%; cognitive data, 66.67%).

The results suggest that FEP subjects are identifiable at the individual level using a range of biological and cognitive measures. Comparatively, only sMRI and DTI allowed discrimination of UHR from HC subjects. For the first time FEP and UHR subjects have been shown to be directly differentiable at the single-subject level using cognitive, sMRI and fMRI data. Preliminarily, the results support clinical development of SVM to help inform identification of FEP and UHR subjects, though future work is needed to provide enhanced levels of accuracy.

The aim of this study was to examine prospective predictors of suicide events, defined as suicide attempts or emergency interventions to reduce suicide risk, in 119 adolescents admitted to an in-patient psychiatric unit for suicidal behaviors and followed naturalistically for 6 months.

Structured diagnostic interviews and self-report instruments were administered to adolescent participants and their parent(s) to assess demographic variables, history of suicidal behavior, psychiatric disorders, family environment and personality/temperament.

Baseline variables that significantly predicted time to a suicide event during follow-up were Black race, high suicidal ideation in the past month, post-traumatic stress disorder (PTSD), childhood sexual abuse (CSA), borderline personality disorder (BPD), low scores on positive affectivity, and high scores on aggression. In a multivariate Cox regression analysis, only Black race, CSA, positive affect intensity and high aggression scores remained significant.

Our findings suggest the following for adolescent populations: (1) in a very high-risk population, risk factors for future attempts may be more difficult to ascertain and some established risk factors (e.g. past suicide attempt) may not distinguish as well; and (2) cross-cutting constructs (e.g. affective and behavioral dysregulation) that underlie multiple psychiatric disorders may be stronger predictors of recurrent suicide events than psychiatric diagnoses. Our finding with respect to positive affect intensity is novel and may have practical implications for the assessment and treatment of adolescent suicide attempters.

To examine whether lifetime DSM-IV diagnosis of major depressive disorder (MDD), including age at onset and number of episodes, is associated with brain atrophy in older persons without dementia.

Within the population-based Age, Gene/Environment Susceptibility (AGES)–Reykjavik Study, 4354 persons (mean age 76 ± 5 years, 58% women) without dementia had a 1.5-T brain magnetic resonance imaging (MRI) scan. Automated brain segmentation total and regional brain volumes were calculated. History of MDD, including age at onset and number of episodes, and MDD in the past 2 weeks was diagnosed according to DSM-IV criteria using the Mini-International Neuropsychiatric Interview (MINI).

Of the total sample, 4.5% reported a lifetime history of MDD; 1.5% had a current diagnosis of MDD (including 75% with a prior history of depression) and 3.0% had a past but no current diagnosis (remission). After adjusting for multiple covariates, compared to participants never depressed, those with current MDD (irrespective of past) had more global brain atrophy [B = –1.25%, 95% confidence interval (CI) −2.05 to −0.44], including more gray- and white-matter atrophy in most lobes, and also more atrophy of the hippocampus and thalamus. Participants with current, first-onset MDD also had more brain atrophy (B = –1.62%, 95% CI −3.30 to 0.05) whereas those remitted did not (B = 0.06%, 95% CI −0.54 to 0.66).

In older persons without dementia, current MDD, irrespective of prior history, but not remitted MDD was associated with widespread gray- and white-matter brain atrophy. Prospective studies should examine whether MDD is a consequence of, or contributes to, brain volume loss and development of dementia.

Many research groups have attempted to predict which individuals with an at-risk mental state (ARMS) for psychosis will later develop a psychotic disorder. However, it is difficult to predict the course and outcome based on individual symptoms scores.

Data from 318 ARMS individuals from two specialized services for ARMS subjects were analysed using latent class cluster analysis (LCCA). The score on the Comprehensive Assessment of At-Risk Mental States (CAARMS) was used to explore the number, size and symptom profiles of latent classes.

LCCA produced four high-risk classes, censored after 2 years of follow-up: class 1 (mild) had the lowest transition risk (4.9%). Subjects in this group had the lowest scores on all the CAARMS items, they were younger, more likely to be students and had the highest Global Assessment of Functioning (GAF) score. Subjects in class 2 (moderate) had a transition risk of 10.9%, scored moderately on all CAARMS items and were more likely to be in employment. Those in class 3 (moderate–severe) had a transition risk of 11.4% and scored moderately severe on the CAARMS. Subjects in class 4 (severe) had the highest transition risk (41.2%), they scored highest on the CAARMS, had the lowest GAF score and were more likely to be unemployed. Overall, class 4 was best distinguished from the other classes on the alogia, avolition/apathy, anhedonia, social isolation and impaired role functioning.

The different classes of symptoms were associated with significant differences in the risk of transition at 2 years of follow-up. Symptomatic clustering predicts prognosis better than individual symptoms.

Facial emotion perception (FEP) is a critical human skill for successful social interaction, and a substantial body of literature suggests that explicit FEP is disrupted in major depressive disorder (MDD). Prior research suggests that weakness in FEP may be an important phenomenon underlying patterns of emotion-processing challenges in MDD and the disproportionate frequency of MDD in women.

Women with (n = 24) and without (n = 22) MDD, equivalent in age and education, completed a FEP task during functional magnetic resonance imaging.

The MDD group exhibited greater extents of frontal, parietal and subcortical activation compared with the control group during FEP. Activation in the inferior frontal gyrus (IFG) appeared shifted from a left >right pattern observed in healthy women to a bilateral pattern in MDD women. The ratio of left to right suprathreshold IFG voxels in healthy controls was nearly 3:1, whereas in the MDD group, there was a greater percentage of suprathreshold IFG voxels bilaterally, with no leftward bias. In MDD, relatively greater activation in right IFG compared with left IFG (ratio score) was present and predicted FEP accuracy (r = 0.56, p < 0.004), with an inverse relationship observed between FEP and subgenual cingulate activation (r = − 0.46, p = 0.02).

This study links, for the first time, disrupted IFG activation laterality and increased subgenual cingulate activation with deficient FEP in women with MDD, providing an avenue for imaging-to-assessment translational applications in MDD.

Dimensional models of co-morbidity have the potential to improve the conceptualization of mental disorders in research and clinical work, yet little is known about how relatively uncommon disorders may fit with more common disorders. The present study estimated the meta-structure of psychopathology in the US general population focusing on the placement of five under-studied disorders sharing features of thought disorder: paranoid, schizoid, avoidant and schizotypal personality disorders, and manic episodes as well as bipolar disorder.

Data were drawn from the National Epidemiologic Survey on Alcohol and Related Conditions, a face-to-face interview of 34 653 non-institutionalized adults in the US general population. The meta-structure of 16 DSM-IV Axis I and Axis II psychiatric disorders, as assessed by the Alcohol Use Disorder and Associated Disabilities Interview Schedule DSM-IV version (AUDADIS-IV), was examined using exploratory and confirmatory factor analysis.

We document an empirically derived thought disorder factor that is a subdomain of the internalizing dimension, characterized by schizoid, paranoid, schizotypal and avoidant personality disorders as well as manic episodes. Manic episodes exhibit notable associations with both the distress subdomain of the internalizing dimension as well as the thought disorder subdomain. The structure was replicated for bipolar disorder (I or II) in place of manic episodes.

As our understanding of psychopathological meta-structure expands, incorporation of disorders characterized by detachment and psychoticism grows increasingly important. Disorders characterized by detachment and psychoticism may be well conceptualized, organized and measured as a subdimension of the internalizing spectrum of disorders. Manic episodes and bipolar disorder exhibit substantial co-morbidity across both distress and thought disorder domains of the internalizing dimension. Clinically, these results underscore the potential utility of conceptualizing patient treatment needs using an approach targeting psychopathological systems underlying meta-structural classification rubrics.

People with psychosis demonstrate impaired response inhibition on the Stop Signal Task (SST). It is less clear if this impairment extends to reflection impulsivity, a form of impulsivity that has been linked to substance use in non-psychotic samples.

We compared 49 patients with first-episode psychosis (FEP) and 30 healthy control participants on two forms of impulsivity measured using the Information Sampling Test (IST) and the SST, along with clinical and IQ assessments. We also compared those patients who used cannabis with those who had either given up or never used.

Patients with FEP had significantly greater impairment in response inhibition but not in reflection impulsivity compared with healthy controls. By contrast, patients who reported current cannabis use demonstrated greater reflection impulsivity than those that had either given up or never used, whereas there were no differences in response inhibition.

These data suggest that abnormal reflection impulsivity is associated with substance use in psychosis but not psychosis itself; the opposite relationship may hold for response inhibition.

Co-morbidity patterns in epidemiological studies of mental illness consistently demonstrate that a latent internalizing factor accounts for co-morbidity patterns among unipolar mood and anxiety disorders, whereas a latent externalizing factor underlies the covariation of substance-use disorders and antisocial behaviors. However, this structure needs to be extended to include a broader range of disorders.

Exploratory and confirmatory factor analyses were used to examine the structure of co-morbidity using data from the Collaborative Psychiatric Epidemiological Surveys (n = 16 233).

In the best-fitting model, eating and bipolar disorders formed subfactors within internalizing, impulse control disorders were indicators of externalizing, and factor-analytically derived personality disorder scales split between internalizing and externalizing.

This was the first large-scale nationally representative study that has included uncommon mental disorders with sufficient power to examine their fit within a structural model of psychopathology. The results of this study have important implications for conceptualizing myriad mental disorders.