Prediction of treatment outcomes is a key step in improving the treatment of major depressive disorder (MDD). The Canadian Biomarker Integration Network in Depression (CAN-BIND) aims to predict antidepressant treatment outcomes through analyses of clinical assessment, neuroimaging, and blood biomarkers.

In the CAN-BIND-1 dataset of 192 adults with MDD and outcomes of treatment with escitalopram, we applied machine learning models in a nested cross-validation framework. Across 210 analyses, we examined combinations of predictive variables from three modalities, measured at baseline and after 2 weeks of treatment, and five machine learning methods with and without feature selection. To optimize the predictors-to-observations ratio, we followed a tiered approach with 134 and 1152 variables in tier 1 and tier 2 respectively.

A combination of baseline tier 1 clinical, neuroimaging, and molecular variables predicted response with a mean balanced accuracy of 0.57 (best model mean 0.62) compared to 0.54 (best model mean 0.61) in single modality models. Adding week 2 predictors improved the prediction of response to a mean balanced accuracy of 0.59 (best model mean 0.66). Adding tier 2 features did not improve prediction.

A combination of clinical, neuroimaging, and molecular data improves the prediction of treatment outcomes over single modality measurement. The addition of measurements from the early stages of treatment adds precision. Present results are limited by lack of external validation. To achieve clinically meaningful prediction, the multimodal measurement should be scaled up to larger samples and the robustness of prediction tested in an external validation dataset.

Improving Access to Psychological Therapies (IAPT) is a primary care therapy service commissioned by England's National Health Service (NHS) for people with unipolar depression and anxiety-related disorders. Its scope does not extend to ‘severe mental illness’, including bipolar disorders (BD), but evidence suggests there is a high BD prevalence in ostensibly unipolar major depressive disorder (uMDD) samples. This study aimed to indicate the prevalence and characteristics of people with BD in a naturalistic cohort of IAPT patients.

371 participants were assessed before initiating therapy. Participants were categorised by indicated diagnoses: BD type-I (BD-I) or type-II (BD-II) as defined using a DSM diagnostic interview, bipolar spectrum (BSp, not meeting diagnostic criteria but exceeding BD screening thresholds), lifetime uMDD or other. Information about psychiatric history and co-morbidities was examined, along with symptoms before and after therapy.

368 patients provided sufficient data to enable classification. 10% of participants were grouped as having BD-I, 20% BD-II, 40% BSp, 25% uMDD and 5% other. BD and uMDD participants had similar demographic characteristics, but patients meeting criteria for BD-I/BD-II had more complex psychiatric presentations. All three ‘bipolar’ groups had particularly high rates of anxiety disorders. IAPT therapy receipt was comparable between groups, as was therapy response (F9704 = 1.113, p = 0.351).

Notwithstanding the possibility that bipolar diathesis was overestimated, findings illustrate a high prevalence of BD in groups of people notionally with uMDD or anxiety. As well as improving the detection of BD, further substantive investigation is required to establish whether individuals affected by BD should be eligible for primary care psychological intervention.

Adolescence is a key developmental period for the emergence of psychiatric disorders. However, there is still no consensus on the core mechanisms of dysfunction in youth. Neurobiological sensitivity to unpredictable threat has been associated with several psychiatric disorders in adults. The present study examined adolescent defensive motivation (startle reflex) and attention (event-related potentials) in anticipation of unpredictable threat in relation to both adolescent and maternal (i.e. familial risk) internalizing and externalizing spectra.

The sample included 395 15-year-old adolescents and their biological mothers. Adolescent startle potentiation and probe P300 suppression (indicating increased attention to threat) were measured in anticipation of predictable and unpredictable threat. Adolescent and maternal lifetime history of psychiatric disorders were assessed via semi-structured diagnostic interviews, and confirmatory factor analysis was used to model internalizing and externalizing spectra.

The adolescent internalizing spectrum was positively associated with adolescent startle potentiation and probe P300 suppression to unpredictable threat. Conversely, the adolescent externalizing spectrum was negatively associated with adolescent P300 suppression to unpredictable threat. The maternal internalizing spectrum was positively associated with adolescent startle potentiation to unpredictable threat and P300 suppression to both predictable and unpredictable threat. The maternal externalizing spectrum was negatively associated with adolescent startle potentiation to unpredictable threat and P300 suppression to both predictable and unpredictable threat. Adolescent and maternal internalizing and externalizing spectra were independently related to adolescent startle potentiation and P300 suppression.

Adolescent neurobiological sensitivity to unpredictable threat is associated with both personal history and familial risk for the internalizing and externalizing spectra.

Preschool psychiatric symptoms significantly increase the risk for long-term negative outcomes. Transdiagnostic hierarchical approaches that capture general (‘p’) and specific psychopathology dimensions are promising for understanding risk and predicting outcomes, but their predictive utility in young children is not well established. We delineated a hierarchical structure of preschool psychopathology dimensions and tested their ability to predict psychiatric disorders and functional impairment in preadolescence.

Data for 1253 preschool children (mean age = 4.17, s.d. = 0.81) were drawn from three longitudinal studies using a similar methodology (one community sample, two psychopathology-enriched samples) and followed up into preadolescence, yielding a large and diverse sample. Exploratory factor models derived a hierarchical structure of general and specific factors using symptoms from the Preschool Age Psychiatric Assessment interview. Longitudinal analyses examined the prospective associations of preschool p and specific factors with preadolescent psychiatric disorders and functional impairment.

A hierarchical dimensional structure with a p factor at the top and up to six specific factors (distress, fear, separation anxiety, social anxiety, inattention-hyperactivity, oppositionality) emerged at preschool age. The p factor predicted all preadolescent disorders (ΔR2 = 0.04–0.15) and functional impairment (ΔR2 = 0.01–0.07) to a significantly greater extent than preschool psychiatric diagnoses and functioning. Specific dimensions provided additional predictive power for the majority of preadolescent outcomes (disorders: ΔR2 = 0.06–0.15; functional impairment: ΔR2 = 0.05–0.12).

Both general and specific dimensions of preschool psychopathology are useful for predicting clinical and functional outcomes almost a decade later. These findings highlight the value of transdiagnostic dimensions for predicting prognosis and as potential targets for early intervention and prevention.

As an integral ingredient of human sociality, prosocial behavior requires learning what acts can benefit or harm others. However, it remains unknown how individuals adjust prosocial learning to avoid punishment or to pursue reward. Given that arginine vasopressin (AVP) is a neuropeptide that has been involved in modulating various social behaviors in mammals, it could be a crucial neurochemical facilitator that supports prosocial learning.

In 50 placebo controls and 54 participants with AVP administration, we examined the modulation of AVP on the prosocial learning characterized by reward and punishment framework, as well as its underlying neurocomputational mechanisms combining computational modeling, event-related potentials and oscillations.

We found a self-bias that individuals learn to avoid punishment asymmetrically more severely than reward-seeking. Importantly, AVP increased behavioral performances and learning rates when making decisions to avoid losses for others and to obtain gains for self. These behavioral effects were underpinned by larger responses of stimulus-preceding negativity (SPN) to anticipation, as well as higher punishment-related feedback-related negativity (FRN) for prosocial learning and reward-related P300 for proself benefits, while FRN and P300 neural processes were integrated into theta (4–7 Hz) oscillation at the outcome evaluation stage.

These results suggest that AVP context-dependently up-regulates altruism for concerning others' losses and reward-seeking for self-oriented benefits. Our findings provide insight into the selectively modulatory roles of AVP in prosocial behaviors depending on learning contexts between proself reward-seeking and prosocial punishment-avoidance.

The impact of the coronavirus disease 2019 (COVID-19) pandemic on mental health is still being unravelled. It is important to identify which individuals are at greatest risk of worsening symptoms. This study aimed to examine changes in depression, anxiety and post-traumatic stress disorder (PTSD) symptoms using prospective and retrospective symptom change assessments, and to find and examine the effect of key risk factors.

Online questionnaires were administered to 34 465 individuals (aged 16 years or above) in April/May 2020 in the UK, recruited from existing cohorts or via social media. Around one-third (n = 12 718) of included participants had prior diagnoses of depression or anxiety and had completed pre-pandemic mental health assessments (between September 2018 and February 2020), allowing prospective investigation of symptom change.

Prospective symptom analyses showed small decreases in depression (PHQ-9: −0.43 points) and anxiety [generalised anxiety disorder scale – 7 items (GAD)-7: −0.33 points] and increases in PTSD (PCL-6: 0.22 points). Conversely, retrospective symptom analyses demonstrated significant large increases (PHQ-9: 2.40; GAD-7 = 1.97), with 55% reported worsening mental health since the beginning of the pandemic on a global change rating. Across both prospective and retrospective measures of symptom change, worsening depression, anxiety and PTSD symptoms were associated with prior mental health diagnoses, female gender, young age and unemployed/student status.

We highlight the effect of prior mental health diagnoses on worsening mental health during the pandemic and confirm previously reported sociodemographic risk factors. Discrepancies between prospective and retrospective measures of changes in mental health may be related to recall bias-related underestimation of prior symptom severity.

Emotion processing deficits have been identified as a critical transdiagnostic factor that facilitates distress after trauma exposure. Limited skills in identifying and labelling emotional states (i.e. alexithymia) may present on the more automated (less conscious) end of the spectrum of emotional awareness and clarity. Individuals with alexithymia tend to exhibit a disconcordance between subjective experience and autonomic activity (e.g. where high levels of subjective emotional intensity are associated with low physiological arousal), which may exacerbate distress. Although there is a robust link between alexithymia and trauma exposure, no work to date has explored whether alexithymia is associated with emotional response disconcordance among trauma-exposed adults.

Using a validated trauma script paradigm, the present study explored the impact of alexithymia on emotion response concordance [skin conductance (Galvanic Skin Response, GSR) and Total Mood Disturbance (TMD)] among 74 trauma-exposed adults recruited via a posttraumatic stress disorder (PTSD) treatment clinic and student research programme.

Unlike posttraumatic symptom severity, age, sex, participant type and mood (which showed no effect on emotion response concordance), alexithymia was associated with heightened emotion response disconcordance between GSR and TMD [F(1, 37) = 8.93, p = 0.006], with low GSR being associated with high TMD. Observed effects of the trauma script were entirely accounted for by the interaction with alexithymia, such that those with alexithymia showed a negligible association between subjective and physiological states.

This finding is paramount as it shows that a large proportion of trauma-exposed adults have a divergent emotion engagement profile.

Major depressive disorder (MDD) is one of the growing human mental health challenges facing the global health care system. In this study, the structural connectivity between symptoms of MDD is explored using two different network modeling approaches.

Data are from ‘the Virginia Adult Twin Study of Psychiatric and Substance Use Disorders (VATSPSUD)’. A cohort of N = 2163 American Caucasian female-female twins was assessed as part of the VATSPSUD study. MDD symptoms were assessed using personal structured clinical interviews. Two network analyses were conducted. First, an undirected network model was estimated to explore the connectivity between the MDD symptoms. Then, using a Bayesian network, we computed a directed acyclic graph (DAG) to investigate possible directional relationships between symptoms.

Based on the results of the undirected network, the depressed mood symptom had the highest centrality value, indicating its importance in the overall network of MDD symptoms. Bayesian network analysis indicated that depressed mood emerged as a plausible driving symptom for activating other symptoms. These results are consistent with DSM-5 guidelines for MDD. Also, somatic weight and appetite symptoms appeared as the strongest connections in both networks.

We discuss how the findings of our study might help future research to detect clinically relevant symptoms and possible directional relationships between MDD symptoms defining major depression episodes, which would help identify potential tailored interventions. This is the first study to investigate the network structure of VATSPSUD data using both undirected and directed network models.

Insomnia symptoms are common during the postpartum period, yet interventions remain scarce. This trial aimed to simultaneously examine the efficacy of cognitive behavioural therapy (CBT) and light dark therapy (LDT), targeting different mechanisms, against treatment-as-usual (TAU), in reducing maternal postpartum insomnia symptoms.

This three-arm randomised controlled trial recruited from the general community in Australia. Nulliparous females 4–12 months postpartum with self-reported insomnia symptoms [Insomnia Severity Index (ISI) scores >7] were included; severe medical/psychiatric conditions were excluded. Participants were randomised 1:1:1 to CBT, LDT, or TAU stratified by ISI (< or ⩾14) and infant age (< or ⩾8 months). Participants and principal investigators were unblinded. Six-week interventions were delivered via digital materials and telephone. The primary outcome was insomnia symptoms (ISI), assessed pre-, midpoint-, post- (primary endpoint), and one-month post-intervention. Analyses were intention-to-treat using latent growth models.

114 participants (CBT = 39, LDT = 36, TAU = 39; Mage = 32.20 ± 4.62 years) were randomised. There were significantly greater reductions in ISI scores in CBT and LDT (effect sizes −2.01 and −1.52 respectively, p < 0.001) from baseline to post-intervention compared to TAU; improvements were maintained at follow-up. Similar effects were observed for self-reported sleep disturbance. There were greater reductions in fatigue in CBT (effect size = 0.85, p < 0.001) but not LDT (p = 0.11) compared to TAU. Changes in sleepiness, depression, and anxiety were non-significant compared to TAU (all p > 0.08). Four participants (11%) in the LDT group reported headaches, dizziness, or nausea; no others reported adverse events.

Therapist-assisted CBT and LDT were feasible during the first postpartum year; data at post-intervention and 1-month follow-up support their safety and efficacy in reducing postpartum insomnia symptoms.

Emergency service workers (ESW) are known to be at increased risk of mental disorders but population-level and longitudinal data regarding their risk of suicide are lacking.

Suicide data for 2001–2017 were extracted from the Australian National Coronial Information Service (NCIS) for two occupational groups: ESW (ambulance personnel, fire-fighters and emergency workers, police officers) and individuals employed in all other occupations. Age-standardised suicide rates were calculated and risk of suicide compared using negative binomial regression modelling.

13 800 suicide cases were identified among employed adults (20–69 years) over the study period. The age-standardised suicide rate across all ESW was 14.3 per 100 000 (95% CI 11.0–17.7) compared to 9.8 per 100 000 (95% CI 9.6–9.9) for other occupations. Significant occupational differences in the method of suicide were identified (p < 0.001). There was no evidence for increased risk of suicide among ESW compared to other occupations once age, gender and year of death were accounted for (RR = 0.99, 95% CI 0.84–1.17; p = 0.95). In contrast, there was a trend for ambulance personnel to be at elevated risk of suicide (RR = 1.41, 95% CI 1.00–2.00, p = 0.053).

Whilst age-standardised suicide rates among ESW are higher than other occupations, emergency service work was not independently associated with an increased risk of suicide, with the exception of an observed trend in ambulance personnel. Despite an increased focus on ESW mental health and wellbeing over the last two decades, there was no evidence that rates of suicide among ESW are changing over time.

Studies have shown that people with internet gaming disorder (IGD) exhibit impaired executive control of gaming cravings; however, the neural mechanisms underlying this process remain unknown. In addition, these conclusions were based on the hypothesis that brain networks are temporally static, neglecting dynamic changes in cognitive processes.

Resting-state fMRI data were collected from 402 subjects [162 subjects with IGD and 240 recreational game users (RGUs)]. The community structure (recruitment and integration) of the executive control network (ECN) and the basal ganglia network (BGN), which represents the reward network, of patients with IGD and RGUs were compared. Mediation effects among the different networks were analyzed.

Compared to RGUs, subjects with IGD had a lower recruitment coefficient within the right ECN. Further analysis showed that only male subjects had a lower recruitment coefficient. Mediation analysis showed that the integration coefficient of the right ECN mediated the relationship between the recruitment coefficients of both the right ECN and the BGN in RGUs.

Male subjects with IGD had a lower recruitment coefficient than RGUs, which impairing their impulse control. The mediation results suggest that top-down executive control of the ECN is absent in subjects with IGD. Together, these findings could explain why subjects with IGD exhibit impaired executive control of gaming cravings; these results have important therapeutic implications for developing effective interventions for IGD.

Repetitive negative thinking (RNT), a cognitive process that encompasses past (rumination) and future (worry) directed thoughts focusing on negative experiences and the self, is a transdiagnostic construct that is especially relevant for major depressive disorder (MDD). Severe RNT often occurs in individuals with severe levels of MDD, which makes it challenging to disambiguate the neural circuitry underlying RNT from depression severity.

We used a propensity score, i.e., a conditional probability of having high RNT given observed covariates to match high and low RNT individuals who are similar in the severity of depression, anxiety, and demographic characteristics. Of 148 MDD individuals, we matched high and low RNT groups (n = 50/group) and used a data-driven whole-brain voxel-to-voxel connectivity pattern analysis to investigate the resting-state functional connectivity differences between the groups.

There was an association between RNT and connectivity in the bilateral superior temporal sulcus (STS), an important region for speech processing including inner speech. High relative to low RNT individuals showed greater connectivity between right STS and bilateral anterior insular cortex (AI), and between bilateral STS and left dorsolateral prefrontal cortex (DLPFC). Greater connectivity in those regions was specifically related to RNT but not to depression severity.

RNT intensity is directly related to connectivity between STS and AI/DLPFC. This might be a mechanism underlying the role of RNT in perceptive, cognitive, speech, and emotional processing. Future investigations will need to determine whether modifying these connectivities could be a treatment target to reduce RNT.

Considerable heterogeneity exists in treatment response to first-line posttraumatic stress disorder (PTSD) treatments, such as Cognitive Processing Therapy (CPT). Relatively little is known about the timing of when during a course of care the treatment response becomes apparent. Novel machine learning methods, especially continuously updating prediction models, have the potential to address these gaps in our understanding of response and optimize PTSD treatment.

Using data from a 3-week (n = 362) CPT-based intensive PTSD treatment program (ITP), we explored three methods for generating continuously updating prediction models to predict endpoint PTSD severity. These included Mixed Effects Bayesian Additive Regression Trees (MixedBART), Mixed Effects Random Forest (MERF) machine learning models, and Linear Mixed Effects models (LMM). Models used baseline and self-reported PTSD symptom severity data collected every other day during treatment. We then validated our findings by examining model performances in a separate, equally established, 2-week CPT-based ITP (n = 108).

Results across approaches were very similar and indicated modest prediction accuracy at baseline (R2 ~ 0.18), with increasing accuracy of predictions of final PTSD severity across program timepoints (e.g. mid-program R2 ~ 0.62). Similar findings were obtained when the models were applied to the 2-week ITP. Neither the MERF nor the MixedBART machine learning approach outperformed LMM prediction, though benefits of each may differ based on the application.

Utilizing continuously updating models in PTSD treatments may be beneficial for clinicians in determining whether an individual is responding, and when this determination can be made.

It remains unclear how SSRIs and other antidepressants are associated with the risk of repeated suicide attempts. We aimed to analyse the association between redeemed antidepressant prescriptions and the risk of repeated suicide attempts, hypothesising that antidepressant treatment is associated with increased risk of repeated suicide attempts.

The study was based on Danish register data and a validated cohort of 1842 suicide attempts. We used three Cox regression models (crude, adjusted and propensity score matched) to analyse the data; these models included both static and dynamic time-dependent factors.

1842 individuals attempted suicide in the study period, with a total of 210 repeated attempts. Individuals redeeming antidepressant prescriptions were more likely to repeat a suicide attempt. All crude models showed all antidepressants to be significant risk factors (HR around 1.39), whereas all adjusted models showed all antidepressants to be insignificant risk factors.

We found no significant increased risk of repeated suicide attempts in individuals redeeming a prescription for any antidepressant (or only SSRIs) when considering the individuals' baseline risk of repetition. This study is based on validated suicide attempts, register data, and strong epidemiology designs, but it still has some limitations, and the results should be replicated and confirmed in other studies.

Major depressive disorder (MDD) was previously associated with negative affective biases. Evidence from larger population-based studies, however, is lacking, including whether biases normalise with remission. We investigated associations between affective bias measures and depressive symptom severity across a large community-based sample, followed by examining differences between remitted individuals and controls.

Participants from Generation Scotland (N = 1109) completed the: (i) Bristol Emotion Recognition Task (BERT), (ii) Face Affective Go/No-go (FAGN), and (iii) Cambridge Gambling Task (CGT). Individuals were classified as MDD-current (n = 43), MDD-remitted (n = 282), or controls (n = 784). Analyses included using affective bias summary measures (primary analyses), followed by detailed emotion/condition analyses of BERT and FAGN (secondary analyses).

For summary measures, the only significant finding was an association between greater symptoms and lower risk adjustment for CGT across the sample (individuals with greater symptoms were less likely to bet more, despite increasingly favourable conditions). This was no longer significant when controlling for non-affective cognition. No differences were found for remitted-MDD v. controls. Detailed analysis of BERT and FAGN indicated subtle negative biases across multiple measures of affective cognition with increasing symptom severity, that were independent of non-effective cognition [e.g. greater tendency to rate faces as angry (BERT), and lower accuracy for happy/neutral conditions (FAGN)]. Results for remitted-MDD were inconsistent.

This suggests the presence of subtle negative affective biases at the level of emotion/condition in association with depressive symptoms across the sample, over and above those accounted for by non-affective cognition, with no evidence for affective biases in remitted individuals.

Relatively few studies have explored the differential contributions of the accumulative dosage of psychotropic medications on mortality in patients with schizophrenia.

We aimed to explore the effects of the exposure dosage of psychotropic medications on mortality during a follow-up period of 5 years with a national cohort of individuals with schizophrenia in 2010. Causes of death were linked through Taiwan's National Mortality Registry. The mean defined daily dose of antipsychotics, antidepressants, mood stabilizers, and sedative-hypnotics, were calculated and survival analyses were conducted.

A total of 102 964 individuals (54 151 men, 52.59%) with schizophrenia were included. Compared to patients with no exposure to antipsychotics, those with antipsychotic exposure had better survival outcomes, regardless of antipsychotic dosage. Antidepressant exposure, in low and moderate dosage, was associated with decreased all-cause mortality; exposure to mood stabilizers appeared to be associated with an increase in all-cause mortality. Although 89.7% of the patients had been prescribed sedative-hypnotics, exposure to sedative-hypnotics was associated with dose-related increased mortality risk [hazard ratio (HR) in low dose group: 1.16, 95% confidence interval (CI) 1.07–1.27; HR in moderate dose: 1.32, 95% CI 1.21–1.44; HR in high dose: 1.83, 95% CI 1.67–2.01)].

The results indicate that in the treatment of schizophrenia, antipsychotics and antidepressants are associated with lower mortality when using adequate dosages and mood stabilizers and sedative-hypnotics with higher mortality compared with no use. Furthermore, exposure to sedative-hypnotics is associated with a dose-related increased mortality risk which warrants clinical attention and further study.

Alcohol binge drinking may compromise the functioning of the nucleus accumbens (NAc), i.e. the neural hub for processing reward and aversive responses.

As socially stressful events pose particular challenges at developmental stages, this research applied the resident–intruder paradigm as a model of social stress, to highlight behavioural neuroendocrine and molecular maladaptive plasticity in rats at withdrawal from binge-like alcohol exposure in adolescence. In search of a rescue agent, cannabidiol (CBD) was selected due to its favourable effects on alcohol- and stress-related harms.

Binge-like alcohol exposed intruder rats displayed a compromised defensive behaviour against the resident and a blunted response of the stress system, in addition to indexes of abnormal dopamine (DA)/glutamate plasticity and dysfunctional spine dynamics in the NAc. CBD administration (60 mg/kg) was able to: (1) increase social exploration in the binge-like alcohol exposed intruder rats, at the expenses of freezing time, and in control rats, which received less aggressive attacks from the resident; (2) reduce corticosterone levels independently on alcohol previous exposure; (3) restore DA transmission and (4) facilitate excitatory postsynaptic strength and remodelling.

Overall, the maladaptive behavioural and synaptic plasticity promoted by the intersection between binge-like alcohol withdrawal and exposure to adverse social stress can be rescued by a CBD détente effect that results in a successful defensive strategy, supported by a functional endocrine and synaptic plasticity. The current data highlight CBD's relevant therapeutic potential in alcohol- and stress-related harms, and prompt further investigation on its molecular targets.

Empirical evidence suggests that people use cannabis to ameliorate anxiety and depressive symptoms, yet cannabis also acutely worsens psychosis and affective symptoms. However, the temporal relationship between cannabis use, anxiety and depressive symptoms and psychotic experiences (PE) in longitudinal studies is unclear. This may be informed by examination of mutually mediating roles of cannabis, anxiety and depressive symptoms in the emergence of PE.

Data were derived from the second longitudinal Netherlands Mental Health Survey and Incidence Study. Mediation analysis was performed to examine the relationship between cannabis use, anxiety/depressive symptoms and PE, using KHB logit in STATA while adjusting for age, sex and education status.

Cannabis use was found to mediate the relationship between preceding anxiety, depressive symptoms and later PE incidence, but the indirect contribution of cannabis use was small (for anxiety: % of total effect attributable to cannabis use = 1.00%; for depression: % of total effect attributable to cannabis use = 1.4%). Interestingly, anxiety and depressive symptoms were found to mediate the relationship between preceding cannabis use and later PE incidence to a greater degree (% of total effect attributable to anxiety = 17%; % of total effect attributable to depression = 37%).

This first longitudinal cohort study examining the mediational relationship between cannabis use, anxiety/depressive symptoms and PE, shows that there is a bidirectional relationship between cannabis use, anxiety/depressive symptoms and PE. However, the contribution of anxiety/depressive symptoms as a mediator was greater than that of cannabis.

Altered white matter brain connectivity has been linked to depression. The aim of this study was to investigate the association of markers of white matter connectivity with prevalence, incidence and course of depressive symptoms.

Markers of white matter connectivity (node degree, clustering coefficient, local efficiency, characteristic path length, and global efficiency) were assessed at baseline by 3 T MRI in the population-based Maastricht Study (n = 4866; mean ± standard deviation age 59.6 ± 8.5 years, 49.0% women; 17 406 person-years of follow-up). Depressive symptoms (9-item Patient Health Questionnaire; PHQ-9) were assessed at baseline and annually over seven years of follow-up. Major depressive disorder (MDD) was assessed with the Mini-International Neuropsychiatric Interview at baseline only. We used negative binominal, logistic and Cox regression analyses, and adjusted for demographic, cardiovascular, and lifestyle risk factors.

A lower global average node degree at baseline was associated with the prevalence and persistence of clinically relevant depressive symptoms [PHQ-9 ⩾ 10; OR (95% confidence interval) per standard deviation = 1.21 (1.05–1.39) and OR = 1.21 (1.02–1.44), respectively], after full adjustment. On the contrary, no associations were found of global average node degree with the MDD at baseline [OR 1.12 (0.94–1.32) nor incidence or remission of clinically relevant depressive symptoms [HR = 1.05 (0.95–1.17) and OR 1.08 (0.83–1.41), respectively]. Other connectivity measures of white matter organization were not associated with depression.

Our findings suggest that fewer white matter connections may contribute to prevalent depressive symptoms and its persistence but not to incident depression. Future studies are needed to replicate our findings.