Major depression episode (MDE) and postpartum depression (PPD) have the same diagnosis criteria, but dissimilarities may be present regarding the frequency and structure of depressive symptoms.

We used data from the IGEDEPP Cohort (France) to examine DSM-5 depressive symptoms in two groups of women: 486 with PPD and 871 with a history of non-perinatal MDE. We compare (i) the frequency of each depressive symptom adjusted for the severity of depression, (ii) the global structure of depressive symptom networks, and (iii) the centrality of each symptom in the two networks.

Women with PPD were significantly more likely to have appetite disturbance, psychomotor symptoms, and fatigue than those with MDE, while sadness, anhedonia, sleep disturbance, and suicidal ideation were significantly less common. There were no significant differences in the global structure of depressive symptoms of MDE and PPD. However, the most central criterion of the MDE network was “Sadness” while it was “Suicidal ideations” for the PPD network. “Sleep” and “Suicidal ideations” criteria were more central for PPD network, whereas “Culpability” was more important for MDE network than for PPD network.

We found differences in depressive symptoms expression between PPD and MDE, which justify continuing to clinically distinguish PPD from MDE.

Studies reporting that highly intelligent individuals have more mental health disorders often have sampling bias, no or inadequate control groups, or insufficient sample size. We addressed these caveats by examining the difference in the prevalence of mental health disorders between individuals with high and average general intelligence (g-factor) in the UK Biobank.

Participants with g-factor scores standardized relative to the same-age UK population, were divided into two groups: a high g-factor group (g-factor 2 SD above the UK mean; N = 16,137) and an average g-factor group (g-factor within 2 SD of the UK mean; N = 236,273). Using self-report questionnaires and medical diagnoses, we examined group differences in the prevalence of 32 phenotypes, including mental health disorders, trauma, allergies, and other traits.

High and average g-factor groups differed across 15/32 phenotypes and did not depend on sex and/or age. Individuals with high g-factors had less general anxiety (odds ratio [OR] = 0.69, 95% CI [0.64;0.74]) and post-traumatic stress disorder (PTSD; OR = 0.67, 95 %CI [0.61;0.74]), were less neurotic (β = −0.12, 95% CI [−0.15;−0.10]), less socially isolated (OR = 0.85, 95% CI [0.80;0.90]), and were less likely to have experienced childhood stressors and abuse, adulthood stressors, or catastrophic trauma (OR = 0.69–0.90). However, they generally had more allergies (e.g., eczema; OR = 1.13–1.33).

The present study provides robust evidence that highly intelligent individuals do not have more mental health disorders than the average population. High intelligence even appears as a protective factor for general anxiety and PTSD.

The Community Assessment of Psychic Experiences (CAPE) is a 42-item self-report questionnaire that has been developed and validated to measure the dimensions of psychosis in the general population. The CAPE has a three-factor structure with dimensions of positive, negative and depression. Assessing the cross-national equivalence of a questionnaire is an essential prerequisite before pooling data from different countries. In this study, our aim was to investigate the measurement invariance of the CAPE across different countries.

Data were drawn from the European Union Gene-Environment Interaction (EU-GEI) study. Participants (incident cases of psychotic disorder, controls and siblings of cases) were recruited in Brazil, France, Italy, the Netherlands, Spain and UK. To analyse the measurement invariance across these samples, we tested configural invariance (i.e. identical structures of the factors), metric invariance (i.e. equivalence of the factor loadings) and scalar invariance (i.e. equivalence of the thresholds) of the three CAPE dimensions using multigroup categorical confirmatory factor analysis methods.

The configural invariance model fits well, providing evidence for identical factorial structure across countries. In comparison with the configural model invariance, the fit indices were very similar in the metric and scalar invariance models, indicating that factor loadings and thresholds did not differ across the six countries.

We found that, across six countries, the CAPE showed equivalent factorial structure, factor loadings and thresholds. Thus, differences observed in scores between individuals from different countries should be considered as reflecting different levels of psychosis.