Among the most disabling and fatal psychiatric illnesses, eating disorders (EDs) often manifest early in life, which encourages investigations into in utero and perinatal environmental risk factors. The objective of this study was to determine whether complications during pregnancy and birth and perinatal conditions are associated with later eating disorder risk in offspring and whether these associations are unique to EDs.

All individuals born in Denmark to Danish-born parents 1989–2010 were included in the study and followed from their 6th birthday until the end of 2016. Exposure to factors related to pregnancy, birth, and perinatal conditions was determined using national registers, as were hospital-based diagnoses of anorexia nervosa (AN), bulimia nervosa, and eating disorder not otherwise specified during follow-up. For comparison, diagnoses of depressive, anxiety, and obsessive-compulsive disorders were also included. Cox regression was used to compare hazards of psychiatric disorders in exposed and unexposed individuals.

1 167 043 individuals were included in the analysis. We found that similar to the comparison disorders, prematurity was associated with increased eating disorder risk. Conversely, patterns of increasing risks of EDs, especially in AN, with increasing parental ages differed from the more U-shaped patterns observed for depressive and anxiety disorders.

Our results suggest that pregnancy and early life are vulnerable developmental periods when exposures may influence offspring mental health, including eating disorder risk, later in life. The results suggest that some events pose more global transdiagnostic risk whereas other patterns, such as increasing parental ages, appear more specific to EDs.

Studies have shown that there are overlapping traits and symptoms between autism and psychosis but no study to date has addressed this association from an epidemiological approach in the adult general population. Furthermore, it is not clear whether autistic traits are associated with specific symptoms of psychosis or with psychosis in general. We assess these associations for the first time by using the Adult Psychiatric Morbidity Survey (APMS) 2007 and the APMS 2014, predicting an association between autistic traits and probable psychosis, and specific associations between autistic traits and paranoia and strange experiences.

Participants (N = 7353 in 2007 and 7500 in 2014) completed the Psychosis Screening Questionnaire (PSQ) and a 20-item version of the Autism Quotient (AQ-20). Binomial logistic regressions were performed using AQ-20 as the independent variable and probable psychosis and specific symptoms as dependent variables.

In the APMS 2007 dataset, significant associations were found between autism traits and probable psychosis, paranoia, thought insertion, and strange experiences. These results were replicated in APMS 2014 but with the additional significant association between autistic traits and hallucinations. Participants in the highest quartile of the AQ-20, compared with the lowest quartile, had an increased risk of probable psychosis of odds ratio (OR) = 15.5 [95% confidence interval (CI) 4.57–52.6] in APMS 2007 and OR = 22.5 (95% CI 7.64–66.3) in APMS 2014.

Autistic traits are strongly associated with probable psychosis and psychotic experiences with the exception of mania. Limitations such as the cross-sectional nature of the study are discussed.

The typical onset of schizophrenia coincides with the maturational peak in cognition; however, for a significant proportion of patients the onset is before age 18 and after age 30 years. While cognitive deficits are considered core features of schizophrenia, few studies have directly examined the impact of age of illness onset on cognition.

The aim of the study was to examine if the effects of age on cognition differ between healthy controls (HCs) and patients with schizophrenia at illness onset. We examined 156 first-episode antipsychotic-naïve patients across a wide age span (12–43 years), and 161 age- and sex-matched HCs. Diagnoses were made according to ICD-10 criteria. Cognition was assessed using the Brief Assessment of Cognition in Schizophrenia (BACS), and IQ was estimated using subtests from the Wechsler adult- or child-intelligence scales. Multivariate analysis of covariance (MANCOVA) was used to examine linear and quadratic effects of age on cognitive scores and interactions by group, including sex and parental socioeconomic status as covariates.

There was a significant overall effect of age on BACS and IQ (p < 0.001). Significant group-by-age interactions for verbal memory (for age-squared, p = 0.009), and digit sequencing (for age, p = 0.01; age-squared, p < 0.001), indicated differential age-related trajectories between patients and HCs.

Cognitive functions showing protracted maturation into adulthood, such as verbal memory and verbal working memory, may be particularly impaired in both early- and late-schizophrenia onset. Our findings indicate a potential interaction between the timing of neurodevelopmental maturation and a possible premature age effect in late-onset schizophrenia.

In clinical research, populations are often selected on the sum-score of diagnostic criteria such as symptoms. Estimating statistical models where a subset of the data is selected based on a function of the analyzed variables introduces Berkson's bias, which presents a potential threat to the validity of findings in the clinical literature. The aim of the present paper is to investigate the effect of Berkson's bias on the performance of the two most commonly used psychological network models: the Gaussian Graphical Model (GGM) for continuous and ordinal data, and the Ising Model for binary data.

In two simulation studies, we test how well the two models recover a true network structure when estimation is based on a subset of the data typically seen in clinical studies. The network is based on a dataset of 2807 patients diagnosed with major depression, and nodes in the network are items from the Hamilton Rating Scale for Depression (HRSD). The simulation studies test different scenarios by varying (1) sample size and (2) the cut-off value of the sum-score which governs the selection of participants.

The results of both studies indicate that higher cut-off values are associated with worse recovery of the network structure. As expected from the Berkson's bias literature, selection reduced recovery rates by inducing negative connections between the items.

Our findings provide evidence that Berkson's bias is a considerable and underappreciated problem in the clinical network literature. Furthermore, we discuss potential solutions to circumvent Berkson's bias and their pitfalls.

Psychosis, and in particular auditory verbal hallucinations (AVHs), are associated with adversity exposure. However, AVHs also occur in populations with no need for care or distress.

This study investigated whether adversity exposure would differentiate clinical and healthy voice-hearers within the context of a ‘three-hit’ model of vulnerability and stress exposure.

Samples of 57 clinical and 45 healthy voice-hearers were compared on the three ‘hits’: familial risk; adversity exposure in childhood and in adolescence/adulthood.

Clinical voice-hearers showed greater familial risk than healthy voice-hearers, with more family members with a history of psychosis, but not with other mental disorders. The two groups did not differ in their exposure to adversity in childhood [sexual and non-sexual, victimisation; discrimination and socio-economic status (SES)]. Contrary to expectations, clinical voice-hearers did not differ from healthy voice-hearers in their exposure to victimisation (sexual/non-sexual) and discrimination in adolescence/adulthood, but reported more cannabis and substance misuse, and lower SES.

The current study found no evidence that clinical and healthy voice-hearers differ in lifetime victimisation exposure, suggesting victimisation may be linked to the emergence of AVHs generally, rather than need-for-care. Familial risk, substance misuse and lower SES may be additional risk factors involved in the emergence of need-for-care and distress.

Depression is highly prevalent and marked by a chronic and recurrent course. Despite being a major cause of disability worldwide, little is known regarding the determinants of its heterogeneous course. Machine learning techniques present an opportunity to develop tools to predict diagnosis and prognosis at an individual level.

We examined baseline (2008–2010) and follow-up (2012–2014) data of the Brazilian Longitudinal Study of Adult Health (ELSA-Brasil), a large occupational cohort study. We implemented an elastic net regularization analysis with a 10-fold cross-validation procedure using socioeconomic and clinical factors as predictors to distinguish at follow-up: (1) depressed from non-depressed participants, (2) participants with incident depression from those who did not develop depression, and (3) participants with chronic (persistent or recurrent) depression from those without depression.

We assessed 15 105 and 13 922 participants at waves 1 and 2, respectively. The elastic net regularization model distinguished outcome levels in the test dataset with an area under the curve of 0.79 (95% CI 0.76–0.82), 0.71 (95% CI 0.66–0.77), 0.90 (95% CI 0.86–0.95) for analyses 1, 2, and 3, respectively.

Diagnosis and prognosis related to depression can be predicted at an individual subject level by integrating low-cost variables, such as demographic and clinical data. Future studies should assess longer follow-up periods and combine biological predictors, such as genetics and blood biomarkers, to build more accurate tools to predict depression course.

The relationship between the subtypes of psychotic experiences (PEs) and common mental health symptoms remains unclear. The current study aims to establish the 12-month prevalence of PEs in a representative sample of community-dwelling Chinese population in Hong Kong and explore the relationship of types of PEs and common mental health symptoms.

This is a population-based two-phase household survey of Chinese population in Hong Kong aged 16–75 (N = 5719) conducted between 2010 and 2013 and a 2-year follow-up study of PEs positive subjects (N = 152). PEs were measured with Psychosis Screening Questionnaire (PSQ) and subjects who endorsed any item on the PSQ without a clinical diagnosis of psychotic disorder were considered as PE-positive. Types of PEs were characterized using a number of PEs (single v. multiple) and latent class analysis. All PE-positive subjects were assessed with common mental health symptoms and suicidal ideations at baseline and 2-year follow-up. PE status was also assessed at 2-year follow-up.

The 12-month prevalence of PEs in Hong Kong was 2.7% with 21.1% had multiple PEs. Three latent classes of PEs were identified: hallucination, paranoia and mixed. Multiple PEs and hallucination latent class of PEs were associated with higher levels of common mental health symptoms. PE persistent rate at 2-year follow-up was 15.1%. Multiple PEs was associated with poorer mental health at 2-year follow-up.

Results highlighted the transient and heterogeneous nature of PEs, and that multiple PEs and hallucination subtype of PEs may be specific indices of poorer common mental health.