Possible age-related differences in risk of completed suicide following non-fatal self-harm remain unexplored. We examined associations between self-harm and completed suicide across age groups of self-harming patients, and whether these associations varied by violent index method, presence of mental disorder, and repeated self-harm.

The design was a cohort study with linked national registers in Sweden. The study population comprised individuals aged ⩾10 years hospitalized during 1990–1999 due to non-fatal self-harm (n = 53 843; 58% females) who were followed for 9–19 years. We computed hazard ratios (HRs) across age groups (age at index self-harm episode), with time to completed suicide as outcome.

The 1-year HR for suicide among younger males (10–19 years) was 14.6 [95% confidence interval (CI) 4.1–51.9] for violent method and 8.4 (95% CI 1.8–40.0) for mental disorder. By contrast, none of the three potential risk factors increased the 1-year risks in the youngest females. Among patients aged ⩾20 years, the 1-year HR for violent method was 4.6 (95% CI 3.8–5.4) for males and 10.4 (95% CI 8.3–13.0) for females. HRs for repeated self-harm during years 2–9 of follow-up were higher in 10- to 19-year-olds (males: HR 4.0, 95% CI 2.0–7.8; females: HR 3.7, 95% CI 2.1–6.5). The ⩾20 years age groups had higher HRs than the youngest, particularly for females and especially within 1 year.

Violent method and mental disorder increase the 1-year suicide risk in young male self-harm patients. Further, violent method increases suicide risk within 1 year in all age and gender groups except the youngest females. Repeated self-harm may increase the long-term risk more in young patients. These aspects should be accounted for in clinical suicide risk assessment.

Suicide in China has declined since the 1990s. However, there has been limited investigation of the potential spatiotemporal variation and social determinants of suicide during subsequent periods.

Annual suicide counts from 2006 to 2012 stratified by county, 5-year age group (⩾15 years) and gender were obtained from the Chinese Disease Surveillance Points system. Trends and geographic differentials were examined using multilevel negative binomial regression models to explore spatiotemporal variation in suicide, and the role of key sociodemographic factors associated with suicide.

The suicide rate (per 100 000) in China decreased from 14.7 to 9.1, 2006–2012. Rates of suicide were higher in males than females and increased substantially with age. Suicide rates were higher in rural areas compared with urban areas; however, urban–rural disparities reduced over time with a faster decline for rural areas. Within both urban and rural areas, higher rates of suicide were evident in areas with lower socio-economic circumstances (SEC) [rate ratio (RR) 1.85, 95% confidence interval (CI) 1.31–2.62]. Suicide rates varied more than twofold (median RR 2.06) across counties, and were highest in central and southwest regions of China. A high proportion of the divorced population, especially for younger females, was associated with lower suicide rates (RR 0.60, 95% CI 0.46–0.79).

Geographic variations for suicide should be taken into account in policy making, particularly for older males living in rural areas and urban areas with low SEC. Measures to reduce disparities in socio-economic level and alleviate family relation stress are current priorities.

There are currently no neuroanatomical biomarkers of anorexia nervosa (AN) available to make clinical inferences at an individual subject level. We present results of a multivariate machine learning (ML) approach utilizing structural neuroanatomical scan data to differentiate AN patients from matched healthy controls at an individual subject level.

Structural neuroimaging scans were acquired from 15 female patients with AN (age = 20, s.d. = 4 years) and 15 demographically matched female controls (age = 22, s.d. = 3 years). Neuroanatomical volumes were extracted using the FreeSurfer software and input into the Least Absolute Shrinkage and Selection Operator (LASSO) multivariate ML algorithm. LASSO was ‘trained’ to identify ‘novel’ individual subjects as either AN patients or healthy controls. Furthermore, the model estimated the probability that an individual subject belonged to the AN group based on an individual scan.

The model correctly predicted 25 out of 30 subjects, translating into 83.3% accuracy (sensitivity 86.7%, specificity 80.0%) (p < 0.001; χ2 test). Six neuroanatomical regions (cerebellum white matter, choroid plexus, putamen, accumbens, the diencephalon and the third ventricle) were found to be relevant in distinguishing individual AN patients from healthy controls. The predicted probabilities showed a linear relationship with drive for thinness clinical scores (r = 0.52, p < 0.005) and with body mass index (BMI) (r = −0.45, p = 0.01).

The model achieved a good predictive accuracy and drive for thinness showed a strong neuroanatomical signature. These results indicate that neuroimaging scans coupled with ML techniques have the potential to provide information at an individual subject level that might be relevant to clinical outcomes.

Depression is a prevalent disorder that significantly affects the social functioning and interpersonal relationships of individuals. This highlights the need for investigation of the neural mechanisms underlying these social difficulties. Investigation of social exchanges has traditionally been challenging as such interactions are difficult to quantify. Recently, however, neuroeconomic approaches that combine multiplayer behavioural economic paradigms and neuroimaging have provided a framework to operationalize and quantify the study of social interactions and the associated neural substrates.

We investigated brain activation using functional magnetic resonance imaging (fMRI) in unmedicated depressed participants (n = 25) and matched healthy controls (n = 25). During scanning, participants played a behavioural economic paradigm, the Ultimatum Game (UG). In this task, participants accept or reject monetary offers from other players.

In comparison to controls, depressed participants reported decreased levels of happiness in response to ‘fair’ offers. With increasing fairness of offers, controls activated the nucleus accumbens and the dorsal caudate, regions that have been reported to process social information and responses to rewards. By contrast, participants with depression failed to activate these regions with increasing fairness, with the lack of nucleus accumbens activation correlating with increased anhedonia symptoms. Depressed participants also showed a diminished response to increasing unfairness of offers in the medial occipital lobe.

Our findings suggest that depressed individuals differ from healthy controls in the neural substrates involved with processing social information. In depression, the nucleus accumbens and dorsal caudate may underlie abnormalities in processing information linked to the fairness and rewarding aspects of other people's decisions.

The purpose of this study was to use selected Cambridge Neuropsychological Test Automated Battery (CANTAB) tests to examine the dimensional structure of cognitive dysfunction in first episode of psychosis (FEP) patients compared with cognition in healthy subjects.

A total of 109 FEP patients and 96 healthy volunteers were administered eight CANTAB tests of cognitive function. Principal components analysis (PCA) was used to estimate dimensionality within the test results. The dimensions identified by the PCA were assumed to reflect underlying cognitive traits. The plausibility of latent factor models was estimated using confirmatory factor analysis (CFA). Multi-group CFA (MGCFA) was used to test for measurement invariance of factors between groups. The nature and severity of cognitive deficits amongst patients as opposed to controls were evaluated using a general linear model.

Amongst subjects PCA identified two underlying cognitive traits: (i) a broad cognitive domain; (ii) attention/memory and executive function domains. Corresponding CFA models were built that fitted data well for both FEP patients and healthy volunteers. As in MGCFA latent variables appeared differently defined in patient and control groups, differences had to be ascribed using subtest scores rather than their aggregates. At subtest score level the patients performed significantly worse than healthy subjects in all comparisons (p < 0.001).

Results of this study demonstrate that the structure of underlying cognitive abilities as measured by a selection of CANTAB tests is not the same for healthy individuals and FEP patients, with patients displaying widespread cognitive impairment.

Genetic variation in the gene encoding ZNF804A, a risk gene for schizophrenia, has been shown to affect brain functional endophenotypes of the disorder, while studies of white matter structure have been inconclusive.

We analysed effects of ZNF804A single nucleotide polymorphism rs1344706 on grey and white matter using voxel-based morphometry (VBM) in high-resolution T1-weighted magnetic resonance imaging scans of 62 schizophrenia patients and 54 matched healthy controls.

We found a significant (p < 0.05, family-wise error corrected for multiple comparisons) interaction effect of diagnostic group x genotype for local grey matter in the left orbitofrontal and right and left lateral temporal cortices, where patients and controls showed diverging effects of genotype. Analysing the groups separately (at p < 0.001, uncorrected), variation in rs1344706 showed effects on brain structure within the schizophrenia patients in several areas including the left and right inferior temporal, right supramarginal/superior temporal, right and left inferior frontal, left frontopolar, right and left dorsolateral/ventrolateral prefrontal cortices, and the right thalamus, as well as effects within the healthy controls in left lateral temporal, right anterior insula and left orbitofrontal cortical areas. We did not find effects of genotype of regional white matter in either of the two cohorts.

Our findings demonstrate effects of ZNF804A genetic variation on brain structure, with diverging regional effects in schizophrenia patients and healthy controls in frontal and temporal brain areas. These effects, however, might be dependent on the impact of other (genetic or non-genetic) disease factors.

To investigate trajectories of cognitive decline in patients with different types of dementia compared to controls in a longitudinal study.

In 199 patients with Alzheimer's disease (AD), 10 with vascular dementia (VaD), 26 with dementia with Lewy bodies (DLB), 20 with behavioural variant frontotemporal dementia (bvFTD), 15 with language variant frontotemporal dementia (lvFTD) and 112 controls we assessed five cognitive domains: memory, language, attention, executive and visuospatial functioning, and global cognition (Mini-Mental State Examination, MMSE). All subjects had at least two neuropsychological assessments (median 2, range 2–7). Neuropsychological data were standardized into z scores using baseline performance of controls as reference. Linear mixed models (LMMs) were used to estimate baseline cognitive functioning and cognitive decline over time for each group, adjusted for age, gender and education.

At baseline, patients with dementia performed worse than controls in all cognitive domains (p < 0.05) except visuospatial functioning, which was only impaired in patients with AD and DLB (p < 0.001). During follow-up, patients with AD declined in all cognitive domains (p < 0.001). DLB showed decline in every cognitive domain except language and global cognition. bvFTD showed rapid decline in memory, language, attention and executive functioning (all p < 0.01) whereas visuospatial functioning remained fairly stable. lvFTD declined mostly in attention and executive functioning (p < 0.01). VaD showed decline in attention and executive functioning.

We show cognitive trajectories of different types of dementia. These estimations of natural disease course have important value for the design of clinical trials as neuropsychological measures are increasingly being used as outcome measures.

The quality of the therapeutic alliance (TA) has been invoked to explain the equal effectiveness of different psychotherapies, but prior research is correlational, and does not address the possibility that individuals who form good alliances may have good outcomes without therapy.

We evaluated the causal effect of TA using instrumental variable (structural equation) modelling on data from a three-arm, randomized controlled trial of 308 people in an acute first or second episode of a non-affective psychosis. The trial compared cognitive behavioural therapy (CBT) over 6 weeks plus routine care (RC) v. supportive counselling (SC) plus RC v. RC alone. We examined the effect of TA, as measured by the client-rated CALPAS, on the primary trial 18-month outcome of symptom severity (PANSS), which was assessed blind to treatment allocation.

Both adjunctive CBT and SC improved 18-month outcomes, compared to RC. We showed that, for both psychological treatments, improving TA improves symptomatic outcome. With a good TA, attending more sessions causes a significantly better outcome on PANSS total score [effect size −2.91, 95% confidence interval (CI) −0.90 to −4.91]. With a poor TA, attending more sessions is detrimental (effect size +7.74, 95% CI +1.03 to +14.45).

This is the first ever demonstration that TA has a causal effect on symptomatic outcome of a psychological treatment, and that poor TA is actively detrimental. These effects may extend to other therapeutic modalities and disorders.

Childhood maltreatment (CM) may modify the relationship between major depressive disorder (MDD) and hippocampal volume reduction. To disentangle the impact of MDD and CM on hippocampal volume we investigated the association between MDD and hippocampal volume in persons with and without a history of CM in two independent cohorts.

We used data of 262 participants from the Netherlands Study of Depression and Anxiety (NESDA) (mean age 37 years, 32% male) and 636 participants from the SMART-Medea study (mean age 61 years, 81% male). In both studies a 12-month diagnosis of MDD and CM were assessed using a diagnostic interview. Hippocampal volume was measured in NESDA using FreeSurfer software on 3-T magnetic resonance (MR) images and in SMART it was manually outlined on 1.5-T MR images. With analysis of covariance adjusted for intracranial volume, age, gender and lifestyle factors we estimated the effects of MDD and CM on hippocampal volume.

In both cohorts CM was not significantly associated with hippocampal volume. After pooling the data MDD was associated with smaller hippocampal volume (B = −138.90 mm3, p = 0.05) and the interaction between MDD and CM reached significance (p = 0.04); in participants with CM, MDD was related to smaller hippocampal volume (NESDA: B = −316.8 mm3, p = 0.02; SMART: B = −407.6, p = 0.046), but not in participants without CM (p > 0.05).

Our study shows that in two independent cohorts, particularly in individuals with CM, a diagnosis of MDD is related to smaller hippocampal volume. Prospective studies are needed to further determine through which mechanism CM may amplify the relationship between MDD and hippocampal volume.

Patients with schizophrenia have substantially reduced subjective well-being (SW) compared to healthy individuals. It has been suggested that diminished SW may be related to deficits in the neural processing of reward but this has not been shown directly. We hypothesized that, in schizophrenia, lower SW would be associated with attenuated reward-related activation in the reward network.

Twenty patients with schizophrenia with a range of SW underwent a functional magnetic resonance imaging (fMRI) reward task. The brain activity underlying reward anticipation and outcome in schizophrenia was examined and compared to that of 12 healthy participants using a full factorial analysis. Region of interest (ROI) analyses of areas within the reward network and whole-brain analyses were conducted to reveal neural correlates of SW.

Reward-related neural activity in schizophrenia was not significantly different from that of healthy participants; however, the patients with schizophrenia showed significantly diminished SW. Both ROI and whole-brain analyses confirmed that SW scores in the patients correlated significantly with activity, specifically in the dorsal anterior cingulate cortex (dACC), during both reward anticipation and reward outcome. This association was not seen in the healthy participants.

In patients with schizophrenia, reduced activation of the dACC during multiple aspects of reward processing is associated with lower SW. As the dACC has been widely linked to coupling of reward and action, and the link to SW is apparent over anticipation and outcome, these findings suggest that SW deficits in schizophrenia may be attributable to reduced integration of environmental rewarding cues, motivated behaviour and reward outcome.

Psychotic phenomena are common in the general population but are excluded from diagnostic criteria for mild to moderate depression and anxiety despite their co-occurrence and shared risk factors. We used item response theory modelling to examine whether the co-occurrence of depressive, anxiety and psychotic phenomena is best explained by: (1) a single underlying factor; (2) two separate, uncorrelated factors; (3) two separate yet linked factors; or (4) two separate domains along with an underlying ‘common mental distress’ (CMD) factor. We defined where, along any latent continuum, the psychopathological items contributed most information.

We performed a secondary analysis of cross-sectional, item-level information from measures of depression, anxiety and psychotic experiences in 6617 participants aged 13 years from the Avon Longitudinal Study of Parents and Children (ALSPAC) birth cohort and 977 participants aged 18 years from the ROOTS schools-based sample. We replicated results from one sample in the other and validated the latent factors against an earlier parental measure of mental state.

In both cohorts depression, anxiety and psychotic items were best represented as a bi-factor model with a single, unitary CMD factor on which psychotic items conveyed information about the more severe end (model 4); residual variation remained for psychotic items. The CMD factor was significantly associated with the prior parental measure.

Psychotic phenomena co-occur with depression and anxiety in teenagers and may be a marker of severity in a single, unitary dimension of CMD. Psychotic phenomena should be routinely included in epidemiological assessments of psychiatric morbidity, otherwise the most severe symptomatology remains unmeasured.

Major depression is associated with abnormalities in reward processing at neural and behavioural levels. Neural abnormalities in reward have been described in young people at familial risk of depression but behavioural changes in reward-based decision making have been less studied in this group.

We studied 63 young people (mean age 18.9 years) with a parent with a diagnosis of major depression but who had never been depressed themselves, that is with a positive family history of depression (the FH+ group). Participants performed the Cambridge Gambling Task (CGT), which provides several measures of decision making including deliberation time, quality of decision making, risk taking, risk adjustment and delay aversion. A control group of 49 age- and gender-matched young people with no history of mood disorder in a first-degree relative undertook the same task.

Both FH+ participants and controls had low and equivalent scores on anxiety and depression self-rating scales. Compared to controls, the FH+ participants showed overall lower risk taking, although like controls they made more risky choices as the odds of a favourable outcome increased. No other measures of decision making differed between the two groups.

Young people at increased familial risk of depression have altered risk taking that is not accounted for by current affective symptomatology. Lowered risk taking might represent an impairment in reward seeking, which is one of several changes in reward-based behaviours seen in acutely depressed patients; however, our findings suggest that decreased reward seeking could be part of a risk endophenotype for depression.

There are inconclusive findings regarding whether danger and loss events differentially predict the onset of anxiety and depression.

A community sample of adolescents and young adults (n = 2304, age 14–24 years at baseline) was prospectively followed up in up to four assessments over 10 years. Incident anxiety and depressive disorders were assessed at each wave using the DSM-IV/M-CIDI. Life events (including danger, loss and respectively mixed events) were assessed at baseline using the Munich Event List (MEL). Logistic regressions were used to reveal associations between event types at baseline and incident disorders at follow-up.

Loss events merely predicted incident ‘pure’ depression [odds ratio (OR) 2.4 per standard deviation, 95% confidence interval (CI) 1.5–3.9, p < 0.001] whereas danger events predicted incident ‘pure’ anxiety (OR 2.3, 95% CI 1.1–4.6, p = 0.023) and ‘pure’ depression (OR 2.5, 95% CI 1.7–3.5, p < 0.001). Mixed events predicted incident ‘pure’ anxiety (OR 2.9, 95% CI 1.5–5.7, p = 0.002), ‘pure’ depression (OR 2.4, 95% CI 1.6–3.4, p < 0.001) and their co-morbidity (OR 3.6, 95% CI 1.8–7.0, p < 0.001).

Our results provide further evidence for differential effects of danger, loss and respectively mixed events on incident anxiety, depression and their co-morbidity. Since most loss events referred to death/separation from significant others, particularly interpersonal loss appears to be highly specific in predicting depression.

That delusional and delusion-prone individuals ‘jump to conclusions’ is one of the most robust and important findings in the literature on delusions. However, although the notion of ‘jumping to conclusions’ (JTC) implies gathering insufficient evidence and reaching premature decisions, previous studies have not investigated whether the evidence gathering of delusion-prone individuals is, in fact, suboptimal. The standard JTC effect is a relative effect but using relative comparisons to substantiate absolute claims is problematic. In this study we investigated whether delusion-prone participants jump to conclusions in both a relative and an absolute sense.

Healthy participants (n = 112) completed an incentivized probabilistic reasoning task in which correct decisions were rewarded and additional information could be requested for a small price. This combination of rewards and costs generated optimal decision points. Participants also completed measures of delusion proneness, intelligence and risk aversion.

Replicating the standard relative finding, we found that delusion proneness significantly predicted task decisions, such that the more delusion prone the participants were, the earlier they decided. This finding was robust when accounting for the effects of risk aversion and intelligence. Importantly, high-delusion-prone participants also decided in advance of an objective rational optimum, gathering fewer data than would have maximized their expected payoff. Surprisingly, we found that even low-delusion-prone participants jumped to conclusions in this absolute sense.

Our findings support and clarify the claim that delusion formation is associated with a tendency to ‘jump to conclusions’. In short, most people jump to conclusions, but more delusion-prone individuals ‘jump further’.

Magnetic resonance imaging (MRI) studies have shown that brain abnormalities in psychosis might be progressive during the first years of illness. We sought to determine whether first-episode psychosis (FEP) subjects show progressive regional grey matter (GM) changes compared with controls, and whether those changes are associated with diagnosis, illness course or antipsychotic (AP) use.

Thirty-two subjects with first-episode schizophrenia-spectrum disorders (FESZ), 24 patients with first-episode affective psychoses (FEAP) and 34 controls recruited using a population-based design underwent structural MRI scanning at baseline and at a 5-year follow-up. Regional GM volumes were assessed with voxel-based morphometry (VBM). Patients were treated at community settings, and about half of them remained mainly untreated.

No significant progressive changes in GM regional volumes were observed in either the FESZ or FEAP group overall. However, FESZ subjects with a non-remitting course showed GM decrements in the left superior temporal gyrus (STG) and insula relative to remitted FESZ subjects. Non-remitted FEAP subjects exhibited a GM decrease in the dorsolateral prefrontal cortex (DLPFC) bilaterally in comparison to remitted FEAP subjects. Among FESZ subjects, AP use was associated with regional GM decrements in the right insula and increments in the cerebellum.

Our results suggest that the progression of brain abnormalities in FEP subjects is restricted to those with a poor outcome and differs between diagnosis subgroups. AP intake is associated with a different pattern of GM reductions over time.

To clarify the role of genetic and environmental risk factors in alcohol use disorders (AUDs), we performed a meta-analysis of twin and adoption studies and explored the impact of sex, assessment method (interview v. hospital/population records), and study design (twin v. adoption study) on heritability estimates.

The literature was searched for all unique twin and adoption studies of AUD and identified 12 twin and five adoption studies. The data were then reconstructed and analyzed using ordinal data full information maximum likelihood in the OpenMx program. Heterogeneity was tested with likelihood ratio tests by equating the parameters across studies.

There was no evidence for heterogeneity by study design, sex or assessment method. The best-fit estimate of the heritability of AUD was 0.49 [95% confidence interval (CI) 0.43–0.53], and the proportion of shared environmental variance was 0.10 (95% CI 0.03–0.16). Estimates of unique environmental proportions of variance differed significantly across studies.

AUD is approximately 50% heritable. The multiple genetically informative studies of this syndrome have produced consistent results that support the validity of this heritability estimate, especially given the different potential methodological weaknesses of twin and adoption designs, and of assessments of AUD based on personal interviews v. official records. We also found evidence for modest shared environmental effects suggesting that environmental factors also contribute to the familial aggregation of AUDs.

Schizophrenia (SZ) is characterized by a broad global cognitive impairment that precedes the onset of the disease. By contrast, some studies suggest that premorbid deficits are absent, or even reversed, in bipolar disorder (BD). However, studies have shown impairments in cognitive functioning after the illness onset in both disorders. The aim of this study was to systematically review and meta-analyze those studies that compared premorbid and/or post-onset global cognitive function between SZ and BD.

We searched Medline (PubMed), EMBASE and PsycINFO for studies where information on cognitive functioning was collected in both SZ and BD within the same study or using the same methods.

Compared to healthy comparison groups, SZ patients showed a significant premorbid cognitive impairment [standardized mean difference (SMD) −0.597, 95% confidence interval (CI) −0.707 to −0.487, p < 0.0001] and a large post-onset impairment (SMD −1.369, 95% CI −1.578 to −1.160, p < 0.0001). We found small significant deficits in premorbid intellectual function in the BD group when this was assessed retrospectively (−0.147, 95% CI −0.238 to −0.056, p = 0.001) but not prospectively (−0.029, 95% CI −0.199 to + 0.142, p = 0.744), and moderate cognitive impairment after onset (SMD −0.623, 95% CI −0.717 to −0.529, p < 0.0001).

SZ is characterized by significant deficits in premorbid intellectual function but the evidence regarding premorbid function in BD is equivocal. After illness onset, patients with both disorders seem to suffer a further decline in cognitive function but the magnitude of the impairment remains greater in SZ than in BD.

Autism spectrum disorders (ASDs) are persistent disabling neurodevelopmental disorders clinically evident from early childhood. For the first time, the burden of ASDs has been estimated for the Global Burden of Disease Study 2010 (GBD 2010). The aims of this study were to develop global and regional prevalence models and estimate the global burden of disease of ASDs.

A systematic review was conducted for epidemiological data (prevalence, incidence, remission and mortality risk) of autistic disorder and other ASDs. Data were pooled using a Bayesian meta-regression approach while adjusting for between-study variance to derive prevalence models. Burden was calculated in terms of years lived with disability (YLDs) and disability-adjusted life-years (DALYs), which are reported here by world region for 1990 and 2010.

In 2010 there were an estimated 52 million cases of ASDs, equating to a prevalence of 7.6 per 1000 or one in 132 persons. After accounting for methodological variations, there was no clear evidence of a change in prevalence for autistic disorder or other ASDs between 1990 and 2010. Worldwide, there was little regional variation in the prevalence of ASDs. Globally, autistic disorders accounted for more than 58 DALYs per 100 000 population and other ASDs accounted for 53 DALYs per 100 000.

ASDs account for substantial health loss across the lifespan. Understanding the burden of ASDs is essential for effective policy making. An accurate epidemiological description of ASDs is needed to inform public health policy and to plan for education, housing and financial support services.

The aims of the study were to determine the prevalence of cardiometabolic risk factors and establish the proportion of people with psychosis meeting criteria for the metabolic syndrome (MetS). The study also aimed to identify the key lifestyle behaviours associated with increased risk of the MetS and to investigate whether the MetS is associated with illness severity and degree of functional impairment.

Baseline data were collected as part of a large randomized controlled trial (IMPaCT RCT). The study took place within community mental health teams in five Mental Health NHS Trusts in urban and rural locations across England. A total of 450 randomly selected out-patients, aged 18–65 years, with an established psychotic illness were recruited. We ascertained the prevalence rates of cardiometabolic risk factors, illness severity and functional impairment and calculated rates of the MetS, using International Diabetes Federation (IDF) and National Cholesterol Education Program Third Adult Treatment Panel criteria.

High rates of cardiometabolic risk factors were found. Nearly all women and most men had waist circumference exceeding the IDF threshold for central obesity. Half the sample was obese (body mass index ≥ 30 kg/m2) and a fifth met the criteria for type 2 diabetes mellitus. Females were more likely to be obese than males (61% v. 42%, p < 0.001). Of the 308 patients with complete laboratory measures, 57% (n = 175) met the IDF criteria for the MetS.

In the UK, the prevalence of cardiometabolic risk factors in individuals with psychotic illnesses is much higher than that observed in national general population studies as well as in most international studies of patients with psychosis.

The Beck Depression Inventory, 2nd edition (BDI-II) is widely used in research on depression. However, the minimal clinically important difference (MCID) is unknown. MCID can be estimated in several ways. Here we take a patient-centred approach, anchoring the change on the BDI-II to the patient's global report of improvement.

We used data collected (n = 1039) from three randomized controlled trials for the management of depression. Improvement on a ‘global rating of change’ question was compared with changes in BDI-II scores using general linear modelling to explore baseline dependency, assessing whether MCID is best measured in absolute terms (i.e. difference) or as percent reduction in scores from baseline (i.e. ratio), and receiver operator characteristics (ROC) to estimate MCID according to the optimal threshold above which individuals report feeling ‘better’.

Improvement in BDI-II scores associated with reporting feeling ‘better’ depended on initial depression severity, and statistical modelling indicated that MCID is best measured on a ratio scale as a percentage reduction of score. We estimated a MCID of a 17.5% reduction in scores from baseline from ROC analyses. The corresponding estimate for individuals with longer duration depression who had not responded to antidepressants was higher at 32%.

MCID on the BDI-II is dependent on baseline severity, is best measured on a ratio scale, and the MCID for treatment-resistant depression is larger than that for more typical depression. This has important implications for clinical trials and practice.