Richard Emsley

Mediation and moderation of treatment effects in randomised controlled trials of complex interventions

Complex intervention trials should be able to answer both pragmatic and explanatory questions in order to test the theories motivating the intervention and help understand the underlying nature of the clinical problem being tested. Key to this is the estimation of direct effects of treatment and indirect effects acting through intermediate variables which are measured post-randomisation. Using psychological treatment trials as an example of complex interventions, we review statistical methods which crucially evaluate both direct and indirect effects in the presence of hidden confounding between mediator and outcome. We review the historical literature on mediation and moderation of treatment effects. We introduce two methods from within the existing causal inference literature, principal stratification and structural mean models, and demonstrate how these can be applied in a mediation context before discussing approaches and assumptions necessary for attaining identifiability of key parameters of the basic causal model. Assuming that there is modification by baseline covariates of the effect of treatment (i.e. randomisation) on the mediator (i.e. covariate by treatment interactions), but no direct effect on the outcome of these treatment by covariate interactions leads to the use of instrumental variable methods. We describe how moderation can occur through post-randomisation variables, and extend the principal stratification approach to multiple group methods with explanatory models nested within the principal strata. We illustrate the new methodology with motivating examples of randomised trials from the mental health literature.

Watch-and-wait approach versus surgical resection after chemoradiotherapy for patients with rectal cancer (the OnCoRe project): a propensity-score matched cohort analysis

BACKGROUND Induction of a clinical complete response with chemoradiotherapy, followed by observation via a watch-and-wait approach, has emerged as a management option for patients with rectal cancer. We aimed to address the shortage of evidence regarding the safety of the watch-and-wait approach by comparing oncological outcomes between patients managed by watch and wait who achieved a clinical complete response and those who had surgical resection (standard care). METHODS Oncological Outcomes after Clinical Complete Response in Patients with Rectal Cancer (OnCoRe) was a propensity-score matched cohort analysis study, that included patients of all ages diagnosed with rectal adenocarcinoma without distant metastases who had received preoperative chemoradiotherapy (45 Gy in 25 daily fractions with concurrent fluoropyrimidine-based chemotherapy) at a tertiary cancer centre in Manchester, UK, between Jan 14, 2011, and April 15, 2013. Patients who had a clinical complete response were offered management with the watch-and-wait approach, and patients who did not have a complete clinical response were offered surgical resection if eligible. We also included patients with a clinical complete response managed by watch and wait between March 10, 2005, and Jan 21, 2015, across three neighbouring UK regional cancer centres, whose details were obtained through a registry. For comparative analyses, we derived one-to-one paired cohorts of watch and wait versus surgical resection using propensity-score matching (including T stage, age, and performance status). The primary endpoint was non-regrowth disease-free survival from the date that chemoradiotherapy was started, and secondary endpoints were overall survival, and colostomy-free survival. We used a conservative p value of less than 0·01 to indicate statistical significance in the comparative analyses. FINDINGS 259 patients were included in our Manchester tertiary cancer centre cohort, 228 of whom underwent surgical resection at referring hospitals and 31 of whom had a clinical complete response, managed by watch and wait. A further 98 patients were added to the watch-and-wait group via the registry. Of the 129 patients managed by watch and wait (median follow-up 33 months [IQR 19-43]), 44 (34%) had local regrowths (3-year actuarial rate 38% [95% CI 30-48]); 36 (88%) of 41 patients with non-metastatic local regrowths were salvaged. In the matched analyses (109 patients in each treatment group), no differences in 3-year non-regrowth disease-free survival were noted between watch and wait and surgical resection (88% [95% CI 75-94] with watch and wait vs 78% [63-87] with surgical resection; time-varying p=0·043). Similarly, no difference in 3-year overall survival was noted (96% [88-98] vs 87% [77-93]; time-varying p=0·024). By contrast, patients managed by watch and wait had significantly better 3-year colostomy-free survival than did those who had surgical resection (74% [95% CI 64-82] vs 47% [37-57]; hazard ratio 0·445 [95% CI 0·31-0·63; p<0·0001), with a 26% (95% CI 13-39) absolute difference in patients who avoided permanent colostomy at 3 years between treatment groups. INTERPRETATION A substantial proportion of patients with rectal cancer managed by watch and wait avoided major surgery and averted permanent colostomy without loss of oncological safety at 3 years. These findings should inform decision making at the outset of chemoradiotherapy. FUNDING Bowel Disease Research Foundation.

Hearing loss and cognition: the role of hearing AIDS, social isolation and depression.

Hearing loss is associated with poor cognitive performance and incident dementia and may contribute to cognitive decline. Treating hearing loss with hearing aids may ameliorate cognitive decline. The purpose of this study was to test whether use of hearing aids was associated with better cognitive performance, and if this relationship was mediated via social isolation and/or depression. Structural equation modelling of associations between hearing loss, cognitive performance, social isolation, depression and hearing aid use was carried out with a subsample of the UK Biobank data set (n = 164,770) of UK adults aged 40 to 69 years who completed a hearing test. Age, sex, general health and socioeconomic status were controlled for as potential confounders. Hearing aid use was associated with better cognition, independently of social isolation and depression. This finding was consistent with the hypothesis that hearing aids may improve cognitive performance, although if hearing aids do have a positive effect on cognition it is not likely to be via reduction of the adverse effects of hearing loss on social isolation or depression. We suggest that any positive effects of hearing aid use on cognition may be via improvement in audibility or associated increases in self-efficacy. Alternatively, positive associations between hearing aid use and cognition may be accounted for by more cognitively able people seeking and using hearing aids. Further research is required to determine the direction of association, if there is any direct causal relationship between hearing aid use and better cognition, and whether hearing aid use results in reduction in rates of cognitive decline measured longitudinally.

Autism and other psychiatric comorbidity in neurofibromatosis type 1: evidence from a population‐based study

Aim  To investigate psychopathology in children with neurofibromatosis type 1 (NF1), particularly the prevalence of autism spectrum disorder (ASD) and attention‐deficit–hyperactivity disorder (ADHD) symptomatology, using a population‐based sampling approach.

Genome-wide association analysis identifies novel loci for chronotype in 100,420 individuals from the UK Biobank.

Our sleep timing preference, or chronotype, is a manifestation of our internal biological clock. Variation in chronotype has been linked to sleep disorders, cognitive and physical performance, and chronic disease. Here we perform a genome-wide association study of self-reported chronotype within the UK Biobank cohort (n=100,420). We identify 12 new genetic loci that implicate known components of the circadian clock machinery and point to previously unstudied genetic variants and candidate genes that might modulate core circadian rhythms or light-sensing pathways. Pathway analyses highlight central nervous and ocular systems and fear-response-related processes. Genetic correlation analysis suggests chronotype shares underlying genetic pathways with schizophrenia, educational attainment and possibly BMI. Further, Mendelian randomization suggests that evening chronotype relates to higher educational attainment. These results not only expand our knowledge of the circadian system in humans but also expose the influence of circadian characteristics over human health and life-history variables such as educational attainment.

Neurofibromatosis type 1 and autism spectrum disorder.

OBJECTIVE: To determine the prevalence of autism spectrum disorder (ASD) in Neurofibromatosis Type 1 (NF1). METHODS: Second-phase population-based epidemiologic study using an allcase NF1 registry in a defined UK 4.1 million population area. A total of 109 (52.7%) of 207 responders from the initial screening phase were grouped by using the parent-rated Social Responsiveness Scale (SRS) as significant ASD (SRS≥76; n = 32), moderate ASD (SRS ≥ 60<76; n = 29), or non-ASD (SRS <60, n = 48). Twenty-three cases from the significant ASD group, 16 from moderate ASD, and 8 from non-ASD (total n = 47), invited proportionately by random selection, were seen for detailed confirmatory ascertainment. Assessments on Autism Diagnostic Interview-Revised, Autism Diagnostic Observation Scale-Generic, and verbal IQ were combined by using standard Collaborative Program for Excellence in Autism criteria into an ASD categorization for each case (ASD, broad ASD with partial features, non-ASD). A preplanned weighted analysis was used to derive prevalence estimates for the whole population. RESULTS: Fourteen (29.5%) of 47 showed ASD, 13 (27.7%) broad ASD, and 20 (42.5%) non-ASD. The ASD/broad ASD group showed male predominance (1.7:1.0), but did not differ significantly from the non-ASD group on IQ, age, socioeconomic status, inheritance, physical severity, or education. The population prevalence estimate is 24.9% ASD (95% confidence interval 13.1%–42.1%) and 20.8% broad ASD (95% confidence interval 10.0%–38.1%); a total of 45.7% showing some ASD spectrum phenotype. CONCLUSIONS: Findings indicate high prevalence of ASD in NF1, with implications for clinical practice and further research into NF1 as a single-gene model for autism.

Autism and other psychiatric comorbidity in NF1 ? evidence from a population based study

Aim  To investigate psychopathology in children with neurofibromatosis type 1 (NF1), particularly the prevalence of autism spectrum disorder (ASD) and attention‐deficit–hyperactivity disorder (ADHD) symptomatology, using a population‐based sampling approach.

Current Paranoid Thinking in Patients With Delusions: The Presence of Cognitive-Affective Biases

Background: There has been renewed interest in the influence of affect on psychosis. Psychological research on persecutory delusions ascribes a prominent role to cognitive processes related to negative affect: anxiety leads to the anticipation of threat within paranoia; depressive negative ideas about the self create a sense of vulnerability in which paranoid thoughts flourish; and self-consciousness enhances feelings of the self as a target. The objective of this study was to examine such affective processes in relation to state paranoia in patients with delusions. Methods: 130 patients with delusions in the context of a nonaffective psychosis diagnosis (predominately schizophrenia) were assessed for contemporaneous levels of persecutory ideation on 5 visual analog scales. Measures were taken of anxiety, depression, threat anticipation, interpretation of ambiguity, self-focus, and negative ideas about the self. Results: Of the patients, 85% report paranoid thinking at testing. Symptoms of anxiety and depression were highly prevalent. Current paranoid thinking was associated with anxiety, depression, greater anticipation of threat events, negative interpretations of ambiguous events, a self-focused cognitive style, and negative ideas about the self. Conclusions: The study provides a clear demonstration that a range of emotion-related cognitive biases, each of which could plausibly maintain delusions, are associated with current paranoid thinking in patients with psychosis. We identified biases both in the contents of cognition and in the processing of information. Links between affect and psychosis are central to the understanding of schizophrenia. We conclude that treatment of emotional dysfunction should lead to reductions in current psychotic experiences.

Cognitive Mechanisms of Change in Delusions: An Experimental Investigation Targeting Reasoning to Effect Change in Paranoia

Background: Given the evidence that reasoning biases contribute to delusional persistence and change, several research groups have made systematic efforts to modify them. The current experiment tested the hypothesis that targeting reasoning biases would result in change in delusions. Methods: One hundred and one participants with current delusions and schizophrenia spectrum psychosis were randomly allocated to a brief computerized reasoning training intervention or to a control condition involving computer-based activities of similar duration. The primary hypotheses tested were that the reasoning training intervention, would improve (1) data gathering and belief flexibility and (2) delusional thinking, specifically paranoia. We then tested whether the changes in paranoia were mediated by changes in data gathering and flexibility, and whether working memory and negative symptoms moderated any intervention effects. Results: On an intention-to-treat analysis, there were significant improvements in state paranoia and reasoning in the experimental compared with the control condition. There was evidence that changes in reasoning mediated changes in paranoia, although this effect fell just outside the conventional level of significance after adjustment for baseline confounders. Working memory and negative symptoms significantly moderated the effects of the intervention on reasoning. Conclusion: The study demonstrated the effectiveness of a brief reasoning intervention in improving both reasoning processes and paranoia. It thereby provides proof-of-concept evidence that reasoning is a promising intermediary target in interventions to ameliorate delusions, and thus supports the value of developing this approach as a longer therapeutic intervention.

Neuropsychological functioning and jumping to conclusions in delusions

Background It has been consistently demonstrated that delusions are related to jumping to conclusions (JTC), a data-gathering bias and potential candidate endophenotype of psychosis. Recent research suggests that JTC may be a marker of treatment response. However, we know little about the factors contributing to the occurrence of this reasoning bias. This study investigated the relationship between JTC and hypothesised deficits in working memory, employing standard well-validated neuropsychological tests, in people with current delusions. Method One hundred and twenty six people with schizophrenia spectrum psychosis and current delusions were assessed for current symptoms, and tested for JTC. We compared performance on tests of working memory in those with the reasoning bias and those without. Results As expected, 30–40% of this sample of people with current delusions showed the JTC bias. There were no differences in premorbid IQ between those with and without the JTC reasoning bias. However, the performance of the JTC group was significantly worse on tests of working memory. Conclusions The JTC data-gathering bias is associated with impairments in working memory. New non-pharmacological interventions for people with delusions, designed to improve data gathering, may benefit from incorporating strategies to overcome deficits in working memory.