Trajectories of depression and generalised anxiety symptoms over the course of cognitive behaviour therapy in primary care: an observational, retrospective cohort

BACKGROUND

Cognitive-behavioural therapy (CBT) has been shown to be an effective treatment for depression and anxiety. However, most research has focused on the sum scores of symptoms. Relatively little is known about how individual symptoms respond.

METHODS

Longitudinal models were used to explore how depression and generalised anxiety symptoms behave over the course of CBT in a retrospective, observational cohort of patients from primary care settings (n = 5306). Logistic mixed models were used to examine the probability of being symptom-free across CBT appointments, using the 9-item Patient Health Questionnaire and the 7-item Generalised Anxiety Disorder scale as measures.

RESULTS

All symptoms improve across CBT treatment. The results suggest that low mood/hopelessness and guilt/worthlessness improved quickest relative to other depressive symptoms, with sleeping problems, appetite changes, and psychomotor retardation/agitation improving relatively slower. Uncontrollable worry and too much worry were the anxiety symptoms that improved fastest; irritability and restlessness improved the slowest.

CONCLUSIONS

This research suggests there is a benefit to examining symptoms rather than sum scores alone. Investigations of symptoms provide the potential for precision psychiatry and may explain some of the heterogeneity observed in clinical outcomes when only sum scores are considered.

Self-processing in relation to emotion and reward processing in depression

BACKGROUND

Depression is characterised by a heightened self-focus, which is believed to be associated with differences in emotion and reward processing. However, the precise relationship between these cognitive domains is not well understood. We examined the role of self-reference in emotion and reward processing, separately and in combination, in relation to depression.

METHODS

Adults experiencing varying levels of depression (n = 144) completed self-report depression measures (PHQ-9, BDI-II). We measured self, emotion and reward processing, separately and in combination, using three cognitive tasks.

RESULTS

When self-processing was measured independently of emotion and reward, in a simple associative learning task, there was little association with depression. However, when self and emotion processing occurred in combination in a self-esteem go/no-go task, depression was associated with an increased positive other bias [b = 3.51, 95% confidence interval (CI) 1.24-5.79]. When the self was processed in relation to emotion and reward, in a social evaluation learning task, depression was associated with reduced positive self-biases (b = 0.11, 95% CI 0.05-0.17).

CONCLUSIONS

Depression was associated with enhanced positive implicit associations with others, and reduced positive learning about the self, culminating in reduced self-favouring biases. However, when self, emotion and reward processing occurred independently there was little evidence of an association with depression. Treatments targeting reduced positive self-biases may provide more sensitive targets for therapeutic intervention and potential biomarkers of treatment responses, allowing the development of more effective interventions.

Personalised psychotherapy in primary care: evaluation of data-driven treatment allocation to cognitive-behavioural therapy versus counselling for depression

BACKGROUND

Various effective psychotherapies exist for the treatment of depression; however, only approximately half of patients recover after treatment. In efforts to improve clinical outcomes, research has focused on personalised psychotherapy - an attempt to match patients to treatments they are most likely to respond to.

AIM

The present research aimed to evaluate the benefit of a data-driven model to support clinical decision-making in differential treatment allocation to cognitive-behavioural therapy versus counselling for depression.

METHOD

The present analysis used electronic healthcare records from primary care psychological therapy services for patients receiving cognitive-behavioural therapy (n = 14 544) and counselling for depression (n = 4725). A linear regression with baseline sociodemographic and clinical characteristics was used to differentially predict post-treatment Patient Health Questionnaire (PHQ-9) scores between the two treatments. The benefit of differential prescription was evaluated in a held-out validation sample.

RESULTS

On average, patients who received their model-indicated optimal treatment saw a greater improvement (by 1.78 PHQ-9 points). This translated into 4-10% more patients achieving clinically meaningful changes. However, for individual patients, the estimated differences in benefits of treatments were small and rarely met the threshold for minimal clinically important differences.

CONCLUSION

Precision prescription of psychotherapy based on sociodemographic and clinical characteristics is unlikely to produce large benefits for individual patients. However, the benefits may be meaningful from an aggregate public health perspective when applied at scale.

Relationship between change in social evaluation learning and mood in early antidepressant treatment: A prospective cohort study in primary care

BACKGROUND

Antidepressants are proposed to work by increasing sensitivity to positive versus negative information. Increasing positive affective learning within social contexts may help remediate negative self-schema. We investigated the association between change in biased learning of social evaluations about the self and others, and mood during early antidepressant treatment.

METHOD

Prospective cohort assessing patients recruited from primary care in South West England at four timepoints over the first 8 weeks of antidepressant treatment (n = 29). At each timepoint, participants completed self-report measures of depression (Beck Depression Inventory II (BDI-II) and Patient Health Questionnaire 9 (PHQ-9)), anxiety (Generalised Anxiety Disorder Questionnaire 7 (GAD-7)), and a computerised task measuring learning of social evaluations about the self, a friend and a stranger.

RESULTS

We did not find evidence that learning about the self was associated with a reduction in PHQ-9 (b = 0.08, 95% CI: -0.05, 0.20, p = 0.239) or BDI-II scores (b = 0.10, 95% CI: -0.18, 0.38, p = 0.469). We found some weak evidence that increased positive learning about the friend was associated with a reduction in BDI-II scores (b = 0.30, 95% CI: -0.02, 0.62, p = 0.069). However, exploratory analyses indicated stronger evidence that increased positive learning about the self (b = 0.18, 95% CI: 0.07, 0.28, p = 0.002) and a friend (b = 0.22, 95% CI: 0.10, 0.35, p = 0.001) was associated with reductions in anxiety.

CONCLUSIONS

Change in social evaluation learning was associated with a reduction in anxiety but not depression. Antidepressants may treat anxiety symptoms by remediating negative affective biases towards socially threatening information directed towards the self and close others. However, our findings are based on exploratory analyses within a small sample without a control group and are therefore at risk of type 1 errors and order effects. Further research with larger samples is required.

Depression screening using a non-verbal self-association task: A machine-learning based pilot study

BACKGROUND

Effective screening is important to combat the raising burden of depression and opens a critical time window for early intervention. Clinical use of non-verbal depression screening is nascent, yet a promising and viable candidate to supplement verbal screening. Differential self- and emotion-processing in depression patients were previously reported by non-verbal behavioural assessments, corroborated by neuroimaging findings of distinct neuroanatomical markers. Thus non-verbal validated brain-behaviour based self-emotion-related assessment data reflect physiological differences and may support individual level screening of depression.

METHODS

In this pilot study (n = 84) we collected two longitudinal sessions of behavioural assessment data in a laboratory setting. Depression was assessed using Beck Depression Inventory II (BDI-II), to explore optimal screening methods with machine-learning, and to establish the validity of adapting a novel behavioural assessment focusing on self and emotions for depression screening.

RESULTS

The best machine-learning model achieved high performance in depression screening, 10-Fold cross-validation (CV) Area Under the receiver operating characteristic Curve (AUC) of 0.90 and balanced accuracy of 0.81, using a Gradient Boosting algorithm. Prospective prediction using a model trained with session 1 data to predict session 2 depression status achieved a 10-Fold CV AUC of 0.77 and balanced accuracy of 0.66. We also identified interpretable behavioural signatures for depression patients based on the best model.

CONCLUSION

The study supports the utility of using behavioural data as a viable and cost-effective solution for depression screening, with a potential wide range of applications in clinical settings.

Beyond online participant crowdsourcing: The benefits and opportunities of big team addiction science

Participant crowdsourcing platforms (e.g., MTurk, Prolific) offer numerous advantages to addiction science, permitting access to hard-to-reach populations and enhancing the feasibility of complex experimental, longitudinal, and intervention studies. Yet these are met with equal concerns about participant nonnaivety, motivation, and careless responding, which if not considered can greatly compromise data quality. In this article, we discuss an alternative crowdsourcing avenue that overcomes these issues whilst presenting its own unique advantages-crowdsourcing researchers through big team science. First, we review several contemporary efforts within psychology (e.g., ManyLabs, Psychological Science Accelerator) and the benefits these would yield if they were more widely implemented in addiction science. We then outline our own consortium-based approach to empirical dissertations: a grassroots initiative that trains students in reproducible big team addiction science. In doing so, we discuss potential challenges and their remedies, as well as providing resources to help addiction researchers develop these initiatives. Through researcher crowdsourcing, together we can answer fundamental scientific questions about substance use and addiction, build a literature that is representative of a diverse population of researchers and participants, and ultimately achieve our goal of promoting better global health. (PsycInfo Database Record (c) 2022 APA, all rights reserved).

How much change is enough? Evidence from a longitudinal study on depression in UK primary care

BACKGROUND

The Patient Health Questionnaire (PHQ-9), the Beck Depression Inventory (BDI-II) and the Generalised Anxiety Disorder Assessment (GAD-7) are widely used in the evaluation of interventions for depression and anxiety. The smallest reduction in depressive symptoms that matter to patients is known as the Minimum Clinically Important Difference (MCID). Little empirical study of the MCID for these scales exists.

METHODS

A prospective cohort of 400 patients in UK primary care were interviewed on four occasions, 2 weeks apart. At each time point, participants completed all three questionnaires and a 'global rating of change' scale (GRS). MCID estimation relied on estimated changes in symptoms according to reported improvement on the GRS scale, stratified by baseline severity on the Clinical Interview Schedule (CIS-R).

RESULTS

For moderate baseline severity, those who reported improvement on the GRS had a reduction of 21% (95% confidence interval (CI) -26.7 to -14.9) on the PHQ-9; 23% (95% CI -27.8 to -18.0) on the BDI-II and 26.8% (95% CI -33.5 to -20.1) on the GAD-7. The corresponding threshold scores below which participants were more likely to report improvement were -1.7, -3.5 and -1.5 points on the PHQ-9, BDI-II and GAD-7, respectively. Patients with milder symptoms require much larger reductions as percentage of their baseline to endorse improvement.

CONCLUSIONS

An MCID representing 20% reduction of scores in these scales, is a useful guide for patients with moderately severe symptoms. If treatment had the same effect on patients irrespective of baseline severity, those with low symptoms are unlikely to notice a benefit.

FUNDING

Funding. National Institute for Health Research.

Inhibitory control of positive and negative information and adolescent depressive symptoms: a population-based cohort study

BACKGROUND

Large population-based cohort studies of neuropsychological factors that characterise or precede depressive symptoms are rare. Most studies use small case-control or cross-sectional designs, which may cause selection bias and cannot test temporality. In a large UK population-based cohort, we investigated cross-sectional and longitudinal associations between inhibitory control of positive and negative information and adolescent depressive symptoms.

METHODS

Cohort study of 2328 UK adolescents who completed an affective go/no-go task at age 18. Depressive symptoms were assessed with the Clinical Interview Schedule Revised (CIS-R) and short Mood and Feeling Questionnaire (sMFQ) at age 18, and with the sMFQ 1 year later (age 19). Analyses were multilevel and traditional linear regressions, before and after adjusting for confounders.

RESULTS

Cross-sectionally, we found little evidence that adolescents with more depressive symptoms made more inhibitory control errors [after adjustments, errors increased by 0.04% per 1 s.d. increase in sMFQ score (95% confidence interval 0.02-0.06)], but this association was not observed for the CIS-R. There was no evidence for an influence of valence. Longitudinally, there was no evidence that reduced inhibitory control was associated with future depressive symptoms.

CONCLUSIONS

Inhibitory control of positive and negative information does not appear to be a marker of current or future depressive symptoms in adolescents and would not be a useful target in interventions to prevent adolescent depression. Our lack of convincing evidence for associations with depressive symptoms suggests that the affective go/no-go task is not a promising candidate for future neuroimaging studies of adolescent depression.

Effective dose 50 method as the minimal clinically important difference: Evidence from depression trials

OBJECTIVE

Previous research on the minimal clinically important difference (MCID) for depression and anxiety is based on population averages. The present study aimed to identify the MCID across the spectrum of baseline severity.

STUDY DESIGN AND SETTINGS

The present analysis used secondary data from 2 randomized controlled trials for depression (n = 1,122) to calibrate the Global Rating of Change with the PHQ-9 and GAD-7. The MCID was defined as a change in scores corresponding to a 50% probability of patients "feeling better", given their baseline severity, referred to as Effective Dose 50 (ED50).

RESULTS

MCID estimates depended on baseline severity and ranged from no change for very mild up to 14 points (52%) on the PHQ-9 and up to 10 points (48%) on the GAD-7 for very high severity. The average MCID estimates were 3.7 points (23%) and 3.3 (28%) for the PHQ-9 and GAD-7 respectively.

CONCLUSION

The ED50 method generates MCID estimates across the spectrum of baseline severity, offering greater precision but at the cost of greater complexity relative to population average estimates. This has important implications for evaluations of treatments and clinical practice where users can use these results to tailor the MCID to specific populations according to baseline severities.

Effects of go/no-go training on food-related action tendencies, liking and choice

Inhibitory control training effects on behaviour (e.g. 'healthier' food choices) can be driven by changes in affective evaluations of trained stimuli, and theoretical models indicate that changes in action tendencies may be a complementary mechanism. In this preregistered study, we investigated the effects of food-specific go/no-go training on action tendencies, liking and impulsive choices in healthy participants. In the training task, energy-dense foods were assigned to one of three conditions: 100% inhibition (no-go), 0% inhibition (go) or 50% inhibition (control). Automatic action tendencies and liking were measured pre- and post-training for each condition. We found that training did not lead to changes in approach bias towards trained foods (go and no-go relative to control), but we warrant caution in interpreting this finding as there are important limitations to consider for the employed approach-avoidance task. There was only anecdotal evidence for an effect on food liking, but there was evidence for contingency learning during training, and participants were on average less likely to choose a no-go food compared to a control food after training. We discuss these findings from both a methodological and theoretical standpoint and propose that the mechanisms of action behind training effects be investigated further.

The early impact of COVID-19 on primary care psychological therapy services: A descriptive time series of electronic healthcare records

BACKGROUND

There are growing concerns about the impact of the COVID-19 pandemic on mental health. With government-imposed restrictions as well as a general burden on healthcare systems, the pandemic has the potential to disrupt the access to, and delivery of, mental healthcare.

METHODS

Electronic healthcare records from primary care psychological therapy services (Improving Access to Psychological Therapy) in England were used to examine changes in access to mental health services and service delivery during early stages of the COVID-19 pandemic. A descriptive time series was conducted using data from five NHS trusts to examine patterns in referrals to services (1st January 2019 to 24th May 2020) and appointments (1st January 2020 to 24th May 2020) taking place.

FINDINGS

The number of patients accessing mental health services dropped by an average of 55% in the early weeks after the March 2020 lockdown was announced, reaching a maximum reduction of 74% in the initial 3 weeks after lockdown in the UK, which gradually recovered to a 28% reduction by May. We found some evidence suggesting changes in the sociodemographic and clinical characteristics of referrals. Despite a reduction in access, the impact on appointments appeared limited with service providers shifting to remote delivery of care.

INTERPRETATION

Services appeared to adapt to provide continuity of care in mental healthcare. However, patients accessing services reduced, potentially placing a future burden on service. Despite the observational nature of the data, the present study can inform the planning of service provision and policy.

FUNDING

AD and TS were funded by Innovate UK (KTP #11,105).

Emotional recognition training modifies neural response to emotional faces but does not improve mood in healthy volunteers with high levels of depressive symptoms

BACKGROUND

There is demand for new, effective and scalable treatments for depression, and development of new forms of cognitive bias modification (CBM) of negative emotional processing biases has been suggested as possible interventions to meet this need.

METHODS

We report two double blind RCTs, in which volunteers with high levels of depressive symptoms (Beck Depression Inventory ii (BDI-ii) > 14) completed a brief course of emotion recognition training (a novel form of CBM using faces) or sham training. In Study 1 (N = 36), participants completed a post-training emotion recognition task whilst undergoing functional magnetic resonance imaging to investigate neural correlates of CBM. In Study 2 (N = 190), measures of mood were assessed post-training, and at 2-week and 6-week follow-up.

RESULTS

In both studies, CBM resulted in an initial change in emotion recognition bias, which (in Study 2) persisted for 6 weeks after the end of training. In Study 1, CBM resulted in increases neural activation to happy faces, with this effect driven by an increase in neural activity in the medial prefrontal cortex and bilateral amygdala. In Study 2, CBM did not lead to a reduction in depressive symptoms on the BDI-ii, or on related measures of mood, motivation and persistence, or depressive interpretation bias at either 2 or 6-week follow-ups.

CONCLUSIONS

CBM of emotion recognition has effects on neural activity that are similar in some respects to those induced by Selective Serotonin Reuptake Inhibitors (SSRI) administration (Study 1), but we find no evidence that this had any later effect on self-reported mood in an analogue sample of non-clinical volunteers with low mood (Study 2).

A Reduced Self-Positive Belief Underpins Greater Sensitivity to Negative Evaluation in Socially Anxious Individuals

Positive self-beliefs are important for well-being, and are influenced by how others evaluate us during social interactions. Mechanistic accounts of self-beliefs have mostly relied on associative learning models. These account for choice behaviour but not for the explicit beliefs that trouble socially anxious patients. Neither do they speak to self-schemas, which underpin vulnerability according to psychological research. Here, we compared belief-based and associative computational models of social-evaluation, in individuals that varied in fear of negative evaluation (FNE), a core symptom of social anxiety. We used a novel analytic approach, 'clinically informed model-fitting', to determine the influence of FNE symptom scores on model parameters. We found that high-FNE participants learn more easily from negative feedback about themselves, manifesting in greater self-negative learning rates. Crucially, we provide evidence that this bias is underpinned by an overall reduced belief about self-positive attributes. The study population could be characterized equally well by belief-based or associative models, however large individual differences in model likelihood indicated that some individuals relied more on an associative (model-free), while others more on a belief-guided strategy. Our findings have therapeutic importance, as positive belief activation may be used to specifically modulate learning.

AUTHOR SUMMARY

Understanding how we form and maintain positive self-beliefs is crucial to understanding how things go awry in disorders such as social anxiety. The loss of positive self-belief in social anxiety, especially in inter-personal contexts, is thought to be related to how we integrate evaluative information that we receive from others. We frame this social information integration as a learning problem and ask how people learn whether someone approves of them or not. We thus elucidate why the decrease in positive evaluations manifests only for the self, but not for an unknown other, given the same information. We investigated the mechanics of this learning using a novel computational modelling approach, comparing models that treat the learning process as series of stimulusresponse associations with models that treat learning as updating of beliefs about the self (or another). We show that both models characterise the process well and that individuals higher in symptoms of social anxiety learn more from negative information specifically about the self. Crucially, we provide evidence that this originates from a reduction in the amount of positive attributes that are activated when the individual is placed in a social evaluative context.

Effect of acute citalopram on self-referential emotional processing and social cognition in healthy volunteers

BACKGROUND

Depression is characterised by negative views of the self. Antidepressant treatment may remediate negative self-schema through increasing processing of positive information about the self. Changes in affective processing during social interactions may increase expression of prosocial behaviours, improving interpersonal communications.

AIMS

To examine whether acute administration of citalopram is associated with an increase in positive affective learning biases about the self and prosocial behaviour.

METHOD

Healthy volunteers (n = 41) were randomised to either an acute 20 mg dose of citalopram or matched placebo in a between-subjects double-blind design. Participants completed computer-based cognitive tasks designed to measure referential affective processing, social cognition and expression of prosocial behaviours.

RESULTS

Participants administered citalopram made more cooperative choices than those administered placebo in a prisoner's dilemma task (β = 20%, 95% CI: 2%, 37%). Exploratory analyses indicated that participants administered citalopram showed a positive bias when learning social evaluations about a friend (β = 4.06, 95% CI: 0.88, 7.24), but not about the self or a stranger. Similarly, exploratory analyses found evidence of increased recall of positive words and reduced recall of negative words about others (β = 2.41, 95% CI: 0.89, 3.93), but not the self, in the citalopram group.

CONCLUSIONS

Participants administered citalopram showed greater prosocial behaviours, increased positive recall and increased positive learning of social evaluations towards others. The increase in positive affective bias and prosocial behaviours towards others may, at least partially, be a mechanism of antidepressant effect. However, we found no evidence that citalopram influenced self-referential processing.

The clinical effectiveness of sertraline in primary care and the role of depression severity and duration (PANDA): a pragmatic, double-blind, placebo-controlled randomised trial

BACKGROUND

Depression is usually managed in primary care, but most antidepressant trials are of patients from secondary care mental health services, with eligibility criteria based on diagnosis and severity of depressive symptoms. Antidepressants are now used in a much wider group of people than in previous regulatory trials. We investigated the clinical effectiveness of sertraline in patients in primary care with depressive symptoms ranging from mild to severe and tested the role of severity and duration in treatment response.

METHODS

The PANDA study was a pragmatic, multicentre, double-blind, placebo-controlled randomised trial of patients from 179 primary care surgeries in four UK cities (Bristol, Liverpool, London, and York). We included patients aged 18 to 74 years who had depressive symptoms of any severity or duration in the past 2 years, where there was clinical uncertainty about the benefit of an antidepressant. This strategy was designed to improve the generalisability of our sample to current use of antidepressants within primary care. Patients were randomly assigned (1:1) with a remote computer-generated code to sertraline or placebo, and were stratified by severity, duration, and site with random block length. Patients received one capsule (sertraline 50 mg or placebo orally) daily for one week then two capsules daily for up to 11 weeks, consistent with evidence on optimal dosages for efficacy and acceptability. The primary outcome was depressive symptoms 6 weeks after randomisation, measured by Patient Health Questionnaire, 9-item version (PHQ-9) scores. Secondary outcomes at 2, 6 and 12 weeks were depressive symptoms and remission (PHQ-9 and Beck Depression Inventory-II), generalised anxiety symptoms (Generalised Anxiety Disorder Assessment 7-item version), mental and physical health-related quality of life (12-item Short-Form Health Survey), and self-reported improvement. All analyses compared groups as randomised (intention-to-treat). The study is registered with EudraCT, 2013-003440-22 (protocol number 13/0413; version 6.1) and ISRCTN, ISRCTN84544741, and is closed to new participants.

FINDINGS

Between Jan 1, 2015, and Aug 31, 2017, we recruited and randomly assigned 655 patients-326 (50%) to sertraline and 329 (50%) to placebo. Two patients in the sertraline group did not complete a substantial proportion of the baseline assessment and were excluded, leaving 653 patients in total. Due to attrition, primary outcome analyses were of 550 patients (266 in the sertraline group and 284 in the placebo group; 85% follow-up that did not differ by treatment allocation). We found no evidence that sertraline led to a clinically meaningful reduction in depressive symptoms at 6 weeks. The mean 6-week PHQ-9 score was 7·98 (SD 5·63) in the sertraline group and 8·76 (5·86) in the placebo group (adjusted proportional difference 0·95, 95% CI 0·85-1·07; p=0·41). However, for secondary outcomes, we found evidence that sertraline led to reduced anxiety symptoms, better mental (but not physical) health-related quality of life, and self-reported improvements in mental health. We observed weak evidence that depressive symptoms were reduced by sertraline at 12 weeks. We recorded seven adverse events-four for sertraline and three for placebo, and adverse events did not differ by treatment allocation. Three adverse events were classified as serious-two in the sertraline group and one in the placebo group. One serious adverse event in the sertraline group was classified as possibly related to study medication.

INTERPRETATION

Sertraline is unlikely to reduce depressive symptoms within 6 weeks in primary care but we observed improvements in anxiety, quality of life, and self-rated mental health, which are likely to be clinically important. Our findings support the prescription of SSRI antidepressants in a wider group of participants than previously thought, including those with mild to moderate symptoms who do not meet diagnostic criteria for depression or generalised anxiety disorder.

FUNDING

National Institute for Health Research.

Variation in recognition of happy and sad facial expressions and self-reported depressive symptom severity: A prospective cohort study

OBJECTIVE

Cognitive theories suggest people with depression interpret self-referential social information negatively. However, it is unclear whether these biases precede or follow depression. We investigated whether facial expression recognition was associated with depressive symptoms cross-sectionally and longitudinally.

METHODS

Prospective cohort study of people who had visited UK primary care in the past year reporting depressive symptoms (n = 509). Depressive symptoms were measured using the Patient Health Questionnaire (PHQ-9) at four time-points, 2 weeks apart. A computerised task assessed happy and sad facial expression recognition at three time-points (n = 505 at time 1). The unbiased hit rate measured ability to recognise emotions accounting for any general tendency to identify the emotion when it was not present.

RESULTS

The sample included the full range of depressive symptom severity, with 45% meeting diagnostic criteria for depression. There was no evidence that happy or sad unbiased hit rates were associated with concurrent or subsequent depressive symptoms. There was weak evidence that, for every additional face incorrectly classified as happy, concurrent PHQ-9 scores reduced by 0.05 of a point (95% CI = -0.10 to 0.002, p = 0.06 after adjustment for confounders). This association was strongest for more ambiguous facial expressions (interaction term p<0.001).

LIMITATIONS

This was an observational study with relatively short follow-up (6 weeks) and small changes in depressive symptoms and emotion recognition. Only 7% of invited patients consented to participate.

CONCLUSIONS

Reduced misclassifications of ambiguous faces as happy could be a state marker of depression, but was not associated with subsequent depressive symptoms. Future research should focus on the interpretation of ambiguous social information.

Grassroots Training for Reproducible Science: A Consortium-Based Approach to the Empirical Dissertation

There is a widely acknowledged need to improve the reliability and efficiency of scientific research to increase the credibility of the published scientific literature and accelerate discovery. Widespread improvement requires a cultural shift in both thinking and practice, and better education will be instrumental to achieve this. Here we argue that education in reproducible science should start at the grassroots. We present our model of consortium-based student projects to train undergraduates in reproducible team science. We discuss how with careful design we have aligned collaboration with the current conventions for individual student assessment. We reflect on our experiences of several years running the GW4 Undergraduate Psychology Consortium offering insights we hope will be of practical use to others wishing to adopt a similar approach. We consider the pedagogical benefits of our approach in equipping students with 21st-century skills. Finally, we reflect on the need to shift incentives to reward to team science in global research and how this applies to the reward structures of student assessment.

Exposure to childhood adversity and deficits in emotion recognition: results from a large, population-based sample

BACKGROUND

Emotion recognition skills are essential for social communication. Deficits in these skills have been implicated in mental disorders. Prior studies of clinical and high-risk samples have consistently shown that children exposed to adversity are more likely than their unexposed peers to have emotion recognition skills deficits. However, only one population-based study has examined this association.

METHODS

We analyzed data from children participating in the Avon Longitudinal Study of Parents and Children, a prospective birth cohort (n = 6,506). We examined the association between eight adversities, assessed repeatedly from birth to age 8 (caregiver physical or emotional abuse; sexual or physical abuse; maternal psychopathology; one adult in the household; family instability; financial stress; parent legal problems; neighborhood disadvantage) and the ability to recognize facial displays of emotion measured using the faces subtest of the Diagnostic Assessment of Non-Verbal Accuracy (DANVA) at age 8.5 years. In addition to examining the role of exposure (vs. nonexposure) to each type of adversity, we also evaluated the role of the timing, duration, and recency of each adversity using a Least Angle Regression variable selection procedure.

RESULTS

Over three-quarters of the sample experienced at least one adversity. We found no evidence to support an association between emotion recognition deficits and previous exposure to adversity, either in terms of total lifetime exposure, timing, duration, or recency, or when stratifying by sex.

CONCLUSIONS

Results from the largest population-based sample suggest that even extreme forms of adversity are unrelated to emotion recognition deficits as measured by the DANVA, suggesting the possible immutability of emotion recognition in the general population. These findings emphasize the importance of population-based studies to generate generalizable results.

Effects of exposure to bodies of different sizes on perception of and satisfaction with own body size: two randomized studies

Body dissatisfaction is prevalent among women and associated with subsequent obesity and eating disorders. Exposure to images of bodies of different sizes has been suggested to change the perception of 'normal' body size in others. We tested whether exposure to different-sized (otherwise identical) bodies changes perception of own and others' body size, satisfaction with body size and amount of chocolate consumed. In Study 1, 90 18-25-year-old women with normal BMI were randomized into one of three groups to complete a 15 min two-back task using photographs of women either of 'normal weight' (Body Mass Index (BMI) 22-23 kg m-2), or altered to appear either under- or over-weight. Study 2 was identical except the 96 participants had high baseline body dissatisfaction and were followed up after 24 h. We also conducted a mega-analysis combining both studies. Participants rated size of others' bodies, own size, and satisfaction with size pre- and post-task. Post-task ratings were compared between groups, adjusting for pre-task ratings. Participants exposed to over- or normal-weight images subsequently perceived others' bodies as smaller, in comparison to those shown underweight bodies (p < 0.001). They also perceived their own bodies as smaller (Study 1, p = 0.073; Study 2, p = 0.018; mega-analysis, p = 0.001), and felt more satisfied with their size (Study 1, p = 0.046; Study 2, p = 0.004; mega-analysis, p = 0.006). There were no differences in chocolate consumption. This study suggests that a move towards using images of women with a BMI in the healthy range in the media may help to reduce body dissatisfaction, and the associated risk of eating disorders.

Logical and Methodological Issues Affecting Genetic Studies of Humans Reported in Top Neuroscience Journals

Genetics and neuroscience are two areas of science that pose particular methodological problems because they involve detecting weak signals (i.e., small effects) in noisy data. In recent years, increasing numbers of studies have attempted to bridge these disciplines by looking for genetic factors associated with individual differences in behavior, cognition, and brain structure or function. However, different methodological approaches to guarding against false positives have evolved in the two disciplines. To explore methodological issues affecting neurogenetic studies, we conducted an in-depth analysis of 30 consecutive articles in 12 top neuroscience journals that reported on genetic associations in nonclinical human samples. It was often difficult to estimate effect sizes in neuroimaging paradigms. Where effect sizes could be calculated, the studies reporting the largest effect sizes tended to have two features: (i) they had the smallest samples and were generally underpowered to detect genetic effects, and (ii) they did not fully correct for multiple comparisons. Furthermore, only a minority of studies used statistical methods for multiple comparisons that took into account correlations between phenotypes or genotypes, and only nine studies included a replication sample or explicitly set out to replicate a prior finding. Finally, presentation of methodological information was not standardized and was often distributed across Methods sections and Supplementary Material, making it challenging to assemble basic information from many studies. Space limits imposed by journals could mean that highly complex statistical methods were described in only a superficial fashion. In summary, methods that have become standard in the genetics literature-stringent statistical standards, use of large samples, and replication of findings-are not always adopted when behavioral, cognitive, or neuroimaging phenotypes are used, leading to an increased risk of false-positive findings. Studies need to correct not just for the number of phenotypes collected but also for the number of genotypes examined, genetic models tested, and subsamples investigated. The field would benefit from more widespread use of methods that take into account correlations between the factors corrected for, such as spectral decomposition, or permutation approaches. Replication should become standard practice; this, together with the need for larger sample sizes, will entail greater emphasis on collaboration between research groups. We conclude with some specific suggestions for standardized reporting in this area.

The Burden of the "False-Negatives" in Clinical Development: Analyses of Current and Alternative Scenarios and Corrective Measures

The “false‐negatives” of clinical development are the effective treatments wrongly determined ineffective. Statistical errors leading to “false‐negatives” are larger than those leading to “false‐positives,” especially in typically underpowered early‐phase trials. In addition, “false‐negatives” are usually eliminated from further testing, thereby limiting the information available on them. We simulated the impact of early‐phase power on economic productivity in three developmental scenarios. Scenario 1, representing the current status quo, assumed 50% statistical power at phase II and 90% at phase III. Scenario 2 assumed increased power (80%), and Scenario 3, increased stringency of alpha (1%) at phase II. Scenario 2 led, on average, to a 60.4% increase in productivity and 52.4% increase in profit. Scenario 3 had no meaningful advantages. Our results suggest that additional costs incurred by increasing the power of phase II studies are offset by the increase in productivity. We discuss the implications of our results and propose corrective measures.

Low statistical power in biomedical science: a review of three human research domains

Studies with low statistical power increase the likelihood that a statistically significant finding represents a false positive result. We conducted a review of meta-analyses of studies investigating the association of biological, environmental or cognitive parameters with neurological, psychiatric and somatic diseases, excluding treatment studies, in order to estimate the average statistical power across these domains. Taking the effect size indicated by a meta-analysis as the best estimate of the likely true effect size, and assuming a threshold for declaring statistical significance of 5%, we found that approximately 50% of studies have statistical power in the 0-10% or 11-20% range, well below the minimum of 80% that is often considered conventional. Studies with low statistical power appear to be common in the biomedical sciences, at least in the specific subject areas captured by our search strategy. However, we also observe evidence that this depends in part on research methodology, with candidate gene studies showing very low average power and studies using cognitive/behavioural measures showing high average power. This warrants further investigation.

A manifesto for reproducible science

Improving the reliability and efficiency of scientific research will increase the credibility of the published scientific literature and accelerate discovery. Here we argue for the adoption of measures to optimize key elements of the scientific process: methods, reporting and dissemination, reproducibility, evaluation and incentives. There is some evidence from both simulations and empirical studies supporting the likely effectiveness of these measures, but their broad adoption by researchers, institutions, funders and journals will require iterative evaluation and improvement. We discuss the goals of these measures, and how they can be implemented, in the hope that this will facilitate action toward improving the transparency, reproducibility and efficiency of scientific research.

Preventing the ends from justifying the means: withholding results to address publication bias in peer-review

The evidence that many of the findings in the published literature may be unreliable is compelling. There is an excess of positive results, often from studies with small sample sizes, or other methodological limitations, and the conspicuous absence of null findings from studies of a similar quality. This distorts the evidence base, leading to false conclusions and undermining scientific progress. Central to this problem is a peer-review system where the decisions of authors, reviewers, and editors are more influenced by impressive results than they are by the validity of the study design. To address this, BMC Psychology is launching a pilot to trial a new 'results-free' peer-review process, whereby editors and reviewers are blinded to the study's results, initially assessing manuscripts on the scientific merits of the rationale and methods alone. The aim is to improve the reliability and quality of published research, by focusing editorial decisions on the rigour of the methods, and preventing impressive ends justifying poor means.

Acute anxiety and social inference: An experimental manipulation with 7.5% carbon dioxide inhalation

BACKGROUND

Positive self-bias is thought to be protective for mental health. We previously found that the degree of positive bias when learning self-referential social evaluation decreases with increasing social anxiety. It is unclear whether this reduction is driven by differences in state or trait anxiety, as both are elevated in social anxiety; therefore, we examined the effects on the state of anxiety induced by the 7.5% carbon dioxide (CO2) inhalation model of generalised anxiety disorder (GAD) on social evaluation learning.

METHODS

For our study, 48 (24 of female gender) healthy volunteers took two inhalations (medical air and 7.5% CO2, counterbalanced) whilst learning social rules (self-like, self-dislike, other-like and other-dislike) in an instrumental social evaluation learning task. We analysed the outcomes (number of positive responses and errors to criterion) using the random effects Poisson regression.

RESULTS

Participants made fewer and more positive responses when breathing 7.5% CO2 in the other-like and other-dislike rules, respectively (gas × condition × rule interaction p = 0.03). Individuals made fewer errors learning self-like than self-dislike, and this positive self-bias was unaffected by CO2. Breathing 7.5% CO2 increased errors, but only in the other-referential rules (gas × condition × rule interaction p = 0.003).

CONCLUSIONS

Positive self-bias (i.e. fewer errors learning self-like than self-dislike) seemed robust to changes in state anxiety. In contrast, learning other-referential evaluation was impaired as state anxiety increased. This suggested that the previously observed variations in self-bias arise due to trait, rather than state, characteristics.

Psychotic Experiences and Working Memory: A Population-Based Study Using Signal-Detection Analysis

Psychotic Experiences (PEs) during adolescence index increased risk for psychotic disorders and schizophrenia in adult life. Working memory (WM) deficits are a core feature of these disorders. Our objective was to examine the relationship between PEs and WM in a general population sample of young people in a case control study. 4744 individuals of age 17–18 from Bristol and surrounding areas (UK) were analyzed in a cross-sectional study nested within the Avon Longitudinal Study of Parents and Children (ALSPAC) birth cohort study. The dependent variable was PEs, assessed using the semi-structured Psychosis-Like Symptom Interview (PLIKSi). The independent variable was performance on a computerized numerical n-back working memory task. Signal-Detection Theory indices, including standardized hits rate, false alarms rate, discriminability index (d’) and response bias (c) from 2-Back and 3-Back tasks were calculated. 3576 and 3527 individuals had complete data for 2-Back and 3-Back respectively. Suspected/definite PEs prevalence was 7.9% (N = 374). Strongest evidence of association was seen between PEs and false alarms on the 2-Back, (odds ratio (OR) = 1.17 [95% confidence intervals (CI) 1.01, 1.35]) and 3-back (OR = 1.35 [1.18, 1.54]) and with c (OR = 1.59 [1.09, 2.34]), and lower d’ (OR = 0.76 [0.65, 0.89]), on the 3-Back. Adjustment for several potential confounders, including general IQ, drug exposure and different psycho-social factors, and subsequent multiple imputation of missing data did not materially alter the results. WM is impaired in young people with PEs in the general population. False alarms, rather than poor accuracy, are more closely related to PEs. Such impairment is consistent with different neuropsychological models of psychosis focusing on signal-to-noise discrimination, probabilistic reasoning and impaired reality monitoring as a basis of psychotic symptoms.

Changing mothers' perception of infant emotion: a pilot study

Cognitive bias modification (CBM) techniques, which experimentally retrain abnormal processing of affective stimuli, are becoming established for various psychiatric disorders. Such techniques have not yet been applied to maternal processing of infant emotion, which is affected by various psychiatric disorders. In a pilot study, mothers of children under 3 years old (n =  2) were recruited and randomly allocated to one of three training exercises, aiming either to increase or decrease their threshold of perceiving distress in a morphed continuum of 15 infant facial images. Differences between pre- and post-training threshold were analysed between and within subjects. Compared to baseline thresholds, the threshold for perceiving infant distress decreased in the lowered threshold group (mean difference -1.7 frames, 95 % confidence intervals (CI) -3.1 to -0.3, p = 0.02), increased in the raised threshold group (1.3 frames, 95 % CI 0.6 to 2.1, p < 0.01) and was unchanged in the control group (0.1 frames, 95 % CI -0.8 to 1.1, p = 0.80). Between-group differences were similarly robust in regression models and were not attenuated by potential confounders. The findings suggest that it is possible to change the threshold at which mothers perceive ambiguous infant faces as distressed, either to increase or decrease sensitivity to distress. This small study was intended to provide proof of concept (i.e. that it is possible to alter a mother's perception of infant distress). Questions remain as to whether the effects persist beyond the immediate experimental session, have an impact on maternal behaviour and could be used in clinical samples to improve maternal sensitivity and child outcomes.

Addressing risk of bias in trials of cognitive behavioral therapy

A recent network meta-analysis by Zhu and colleagues reported in the Shanghai Archives of Psychiatry compared two different comparators (psychological placebo and waitlist control) in trials assessing the effectiveness of cognitive behavioral therapy (CBT) for the treatment of generalized anxiety disorder (GAD). CBT was superior to both of these control conditions, but psychological placebo was superior to waitlist. However, we argue that the term ‘psychological placebo’ is a misnomer because the impossibility of effectively blinding participants to treatment allocation in CBT trials makes it impossible to control for placebo effects. This failure to blind participants and therapists – and the resultant high risk of bias – was the main reason Zhu and colleagues found that the overall quality of the evidence supporting the conclusion that CBT is effective for GAD is poor. This is a general problem in all psychotherapy trials, which suffer from well-documented methodological and conceptual problems that prevent adequate placebo control and undermine casual inference. We discuss these problems and suggest potential solutions. We conclude that, while it may be difficult to remove potential bias in randomized controlled trials of psychotherapy, we can improve on the status quo by integrating basic science within applied trials to adjust for these biases and, thus, improve the strength of the causal inferences.

Fear of negative evaluation biases social evaluation inference: evidence from a probabilistic learning task

BACKGROUND

Fear of negative evaluation (FNE) defines social anxiety yet the process of inferring social evaluation, and its potential role in maintaining social anxiety, is poorly understood. We developed an instrumental learning task to model social evaluation learning, predicting that FNE would specifically bias learning about the self but not others.

METHODS

During six test blocks (3 self-referential, 3 other-referential), participants (n = 100) met six personas and selected a word from a positive/negative pair to finish their social evaluation sentences "I think [you are / George is]…". Feedback contingencies corresponded to 3 rules, liked, neutral and disliked, with P[positive word correct] = 0.8, 0.5 and 0.2, respectively.

RESULTS

As FNE increased participants selected fewer positive words (β = -0.4, 95% CI -0.7, -0.2, p = 0.001), which was strongest in the self-referential condition (FNE × condition 0.28, 95% CI 0.01, 0.54, p = 0.04), and the neutral and dislike rules (FNE × condition × rule, p = 0.07). At low FNE the proportion of positive words selected for self-neutral and self-disliked greatly exceeded the feedback contingency, indicating poor learning, which improved as FNE increased.

CONCLUSIONS

FNE is associated with differences in processing social-evaluative information specifically about the self. At low FNE this manifests as insensitivity to learning negative self-referential evaluation. High FNE individuals are equally sensitive to learning positive or negative evaluation, which although objectively more accurate, may have detrimental effects on mental health.

Moderators of response to cognitive behavioural therapy as an adjunct to pharmacotherapy for treatment-resistant depression in primary care

BACKGROUND

Stratified medicine aims to improve clinical and cost-effectiveness by identifying moderators of treatment that indicate differential response to treatment. Cognitive behavioural therapy (CBT) is often offered as a 'next-step' for patients who have not responded to antidepressants, but no research has examined moderators of response to CBT in this population. We aimed, therefore, to identify moderators of response to CBT in treatment resistant depression.

METHODS

We used linear regression to test for interactions between treatment effect and 14 putative moderator variables using data from the CoBalT randomised controlled trial. This trial examined the effectiveness of CBT given in addition to usual care (n=234) compared with usual care alone (n=235) for primary care patients with treatment resistant depression.

RESULTS

Age was the only variable with evidence for effect modification (p Value for interaction term=0.012), with older patients benefiting the most from CBT. We found no evidence of effect modification by any other demographic, life, illness, personality trait, or cognitive variable (p≥0.2).

CONCLUSIONS

Given the largely null findings, a stratified approach that might limit offering CBT is premature; CBT should be offered to all individuals where antidepressant medication has failed.

Are multiple physical symptoms a poor prognostic factor or just a marker of depression severity? Secondary analysis of the GenPod trial

BACKGROUND

Using data from the GenPod trial this study investigates: (i) if depressed individuals with multiple physical symptoms have a poorer response to antidepressants before and after adjustment for baseline Beck Depression Inventory II (BDI-II); and (ii) if reboxetine is more effective than citalopram in depression with multiple physical symptoms.

METHODS

Linear regression models were used to estimate differences in mean BDI-II score at 6 and 12 weeks.

RESULTS

Before adjusting for baseline BDI-II, the difference in mean BDI-II score between no and multiple physical symptoms was 4.5 (95% CI 1.87, 7.14) at 6 weeks, 4.51 (95% CI 1.60, 7.42) at 12 weeks. After adjustment for baseline BDI-II, there was no evidence of a difference in outcome according to physical symptoms with a difference in mean BDI-II of 2.17 (95% CI -0.39, 4.73) at 6 weeks and 2.43 (95% CI -0.46, 5.32) at 12 weeks. There was no evidence that reboxetine was more effective than citalopram in those with multiple physical symptoms at 6 (P=0.18) or 12 weeks (P=0.24).

LIMITATIONS

Differential non-adherence between treatment arms has the potential to bias estimates of treatment efficacy.

CONCLUSION

Multiple physical symptoms predict response to antidepressants, but not after adjustment for baseline depression severity. Physical symptoms could be a marker of severe depression rather than an independent prognostic factor and depression should be considered in patients with multiple physical symptoms. Treatment with reboxetine conferred no advantage over citalopram in those with physical symptoms, and it is less well tolerated.

Peer victimization during adolescence and risk for anxiety disorders in adulthood: a prospective cohort study

BACKGROUND

Peer victimization is ubiquitous across schools and cultures, and has been suggested as one developmental pathway to anxiety disorders. However, there is a dearth of prospective studies examining this relationship. The purpose of this cohort study was to examine the association between peer victimization during adolescence and subsequent anxiety diagnoses in adulthood. A secondary aim was to investigate whether victimization increases risk for severe anxiety presentations involving diagnostic comorbidity.

METHODS

The sample comprised 6,208 adolescents from the Avon Longitudinal Study of Parents and Children who were interviewed about experiences of peer victimization at age 13. Maternal report of her child's victimization was also assessed. Anxiety disorders at age 18 were assessed with the Clinical Interview Schedule-Revised. Multivariable logistic regression was used to examine the association between victimization and anxiety diagnoses adjusted for potentially confounding individual and family factors. Sensitivity analyses explored whether the association was independent of diagnostic comorbidity with depression.

RESULTS

Frequently victimized adolescents were two to three times more likely to develop an anxiety disorder than nonvictimized adolescents (OR = 2.49, 95% CI: 1.62-3.85). The association remained after adjustment for potentially confounding individual and family factors, and was not attributable to diagnostic overlap with depression. Frequently victimized adolescents were also more likely to develop multiple internalizing diagnoses in adulthood.

CONCLUSIONS

Victimized adolescents are at increased risk of anxiety disorders in later life. Interventions to reduce peer victimization and provide support for victims may be an effective strategy for reducing the burden associated with these disorders.

Power failure: why small sample size undermines the reliability of neuroscience

A study with low statistical power has a reduced chance of detecting a true effect, but it is less well appreciated that low power also reduces the likelihood that a statistically significant result reflects a true effect. Here, we show that the average statistical power of studies in the neurosciences is very low. The consequences of this include overestimates of effect size and low reproducibility of results. There are also ethical dimensions to this problem, as unreliable research is inefficient and wasteful. Improving reproducibility in neuroscience is a key priority and requires attention to well-established but often ignored methodological principles.

Social inference and social anxiety: evidence of a fear-congruent self-referential learning bias

BACKGROUND AND OBJECTIVES

Fears of negative evaluation characterise social anxiety, and preferential processing of fear-relevant information is implicated in maintaining symptoms. Little is known, however, about the relationship between social anxiety and the process of inferring negative evaluation. The ability to use social information to learn what others think about one, referred to here as self-referential learning, is fundamental for effective social interaction. The aim of this research was to examine whether social anxiety is associated with self-referential learning.

METHODS

102 Females with either high (n = 52) or low (n = 50) self-reported social anxiety completed a novel probabilistic social learning task. Using trial and error, the task required participants to learn two self-referential rules, 'I am liked' and 'I am disliked'.

RESULTS

Participants across the sample were better at learning the positive rule 'I am liked' than the negative rule 'I am disliked', β = -6.4, 95% CI [-8.0, -4.7], p < 0.001. This preference for learning positive self-referential information was strongest in the lowest socially anxious and was abolished in the most symptomatic participants. Relative to the low group, the high anxiety group were better at learning they were disliked and worse at learning they were liked, social anxiety by rule interaction β = 3.6; 95% CI [+0.3, +7.0], p = 0.03.

LIMITATIONS

The specificity of the results to self-referential processing requires further research.

CONCLUSIONS

Healthy individuals show a robust preference for learning that they are liked relative to disliked. This positive self-referential bias is reduced in social anxiety in a way that would be expected to exacerbate anxiety symptoms.

Factors associated with differential response to online cognitive behavioural therapy

PURPOSE

Patients differ in their response to treatments. There is obvious clinical utility in establishing patient characteristics that are associated with differential treatment responses (i.e., are effect modifiers or moderators of treatment response). Factors that moderate response to cognitive behavioural therapy (CBT) remain unclear. This study investigates whether factors prognostic of general depression outcome generally are also moderators of response to online CBT in a sample of depressed patients recruited through U.K. general practices.

METHODS

Secondary analysis of a randomised controlled trial, internet-based psychotherapy for depression. A total of 297 patients referred from 55 U.K. general practices and suffering from depression were randomly allocated to receive either online CBT or waiting list control. Treatment effect was measured by comparing depression score at 4 months between randomization groups. Treatment effect modification was assessed using regression analyses focusing on interactions between treatment effect and putative moderator variables.

RESULTS

Pretreatment severity and marital status moderated treatment response. More severe patients, and patients who were separated, widowed, or divorced, benefited most from the intervention. Weak evidence suggested that treatment effectiveness diminished with increasing recent adverse life events. No evidence was found to suggest that educational attainment, age, and history of depression-moderated treatment response.

CONCLUSIONS

Secondary analyses of trials comparing two or more treatments allow factors that may moderate treatment response to be distinguished from more general prognostic indicators, although caution is needed in interpreting such exploratory analyses.