Epidemiology in Schizophrenia Research: Moving from Trait to State of the Art

Image result for random news cartoonThis weekend marks the beginning of the 16th International Congress on Schizophrenia Research, the biennial international meeting this year taking place in the pleasant climes of San Diego, US. The meeting offers a chance to hear about groundbreaking research across a range of domains, including cell biology, animal models, cognitive neuroscience and psychiatric genetics. Moreover, there is a collaborative, friendly atmosphere, with junior and senior colleagues engaging in meaningful, critical discourse in and out of the formal sessions.

Despite the warm reception, the inclusion of epidemiological research in the Congress’ program has traditionally been cool. The 16th meeting will not reverse that trend, with one session devoted specifically to epidemiological research, with 10 talks crammed into a mammoth two and a half hour session on Saturday afternoon. To an outsider – or one of the forty or so people attending the sparsely populated ballroom – there will have been little to convince the field, or future programme committees, that epidemiology should move from a mild trait of schizophrenia research to a discipline which should be emblematic of the state of all science conducted within psychiatric and schizophrenia research.

Of the ten scheduled talks, two speakers did not show up, apparently with no prior apologies to the chair. Although there might have been genuine, unavoidable reasons for these cancellations, the lack of courtesy shown to the audience, other speakers, chair and meeting organisers manifests itself as apathetic; a trait that will rarely endear the discipline to future conference committees. To get an oral presentation accepted at a major international meeting should be viewed as an honour – if in doubt, one needs to look no further than the thousands of wonderful poster presentations at ICOSR, of whom I suspect any one would be delighted to receive a wider audience for their research.

Perhaps most concerningly, the quality of the epidemiological research on show was not able to shine through with all the wonderful, disciplined and principled methods that good epidemiological training should permit. In the age of big data, computational psychiatry and causal inference, the session was pockmarked by a series of poorly controlled studies, negative findings (of themselves potentially important, if we can establish that the findings are likely to be true, having considered the caveats of possible chance, bias and confounding) and problematic interpretation of results and their translational meaning for public mental health. There were some highlights in the session, notably some interesting work on polygenetic risk scores which appear not to confound associations between urban living and later psychosis risk, despite people in urban areas having slightly higher PRS for schizophrenia.

As a field, those of us in psychiatric epidemiology have a duty to do more. We need to engage with modern, state of the science with regard to causal inference and other techniques which potentially allow observational epidemiology to inform psychiatric research about aetiological mechanisms in the same way that experimental research has the potential to. Before that, good basic training in epidemiology can serve the whole field of schizophrenia research. Some of psychiatry’s greatest thinkers come from a background in epidemiological training and reasoning. Our core business as epidemiologists is not to show associations between risk factor X and outcome Y, but to bring to bear the principles of designing unbiased, unfounded and robust studies to answer clear research questions with carefully formulated hypotheses. On our MSc in Clinical Mental Health Sciences at UCL, which attracts 80 budding psychologists each year, the epidemiological training our students receive does not seek to teach them how to conduct an epidemiological study, but how to apply critical thinking and epidemiological reasoning to any research “evidence” they may encounter in their future careers. In this respect, our discipline should be proud; we can take a lead in research design across all areas of schizophrenia research, whether in the cell, the mouse or the human. Never has there been a more exciting time to work in mental health research; converging data from social neuroscience, neuropsychology, psychiatric genetics and epidemiology suggest a mix of genetic and environmental factors are critical to shaping psychosis vulnerability. Epidemiologists need to be central to those conversations.

Fortunately, great epidemiological research in schizophrenia does exist, addressing fundamental questions of variance in incidence, prevalence and causes worldwide. For example, forthcoming data from the international multi-site European Network of National Schizophrenia Networks Studying Gene Environment Interactions (EU-GEI) study has brought together a multidisciplinary team of researchers worldwide to test vital questions about how genetic and environmental factors may combine to increase psychosis risk. Such major consortia are never undertaken lightly, and the fruits of this study – amongst many other carefully designed epidemiological studies – should come to bear in future meetings. Until then, let’s promote and celebrate the uses of epidemiological reasoning throughout schizophrenia research, and be part of the conversation about the strength of evidence across the state of science in our field.

Are cats bad for your mental health? Probably, not.

Over the past few years, a number of scientific studies and media outlets have reported that Toxoplasma Gondii (T. Gondii) infection could increase a person’s risk of experiencing adverse mental health outcomes – including schizophrenia, suicide, and intermittent rage disorder.

Since domestic cats are the primary hosts of T. Gondii (i.e., they provide an environment within which this parasite can reproduce) some people have speculated that cat ownership may put people at increased risk of mental illness, by exposing them to T. Gondii infection. While a handful of small studies have found evidence to support a link between cat ownership and psychosis, most of these investigations had serious methodological limitations (e.g.: relying on recall of past cat ownership, small sample sizes, missing data, lack of control for other explanations), which may have biased their findings.

To tackle these limitations, we used data from a large, longitudinal sample of approximately 5,000 children, a sub-group of 14,000 children born in the former region of Avon, (UK) between 1991 and 1992. Unlike previous studies, we were able to follow people over time, from birth to late adolescence, and address a number of the limitations of previous studies, including controlling for alternative explanations (including socioeconomic status, ethnicity, other pet ownership and over-crowding) and taking into account missing data. We studied whether mothers who owned a cat: 1) while pregnant; 2) when the child was 4 years old; and 3) 10 years old, were more likely to have children who reported psychotic symptoms (e.g.: experiences of visual or auditory hallucinations, and of paranoia) in early (age 13) and late adolescence (age 18).

We found that children who were born and raised in households that included cats at any time period were not at a higher risk of having psychotic symptoms when they were 13 or 18 years old. This finding in a large, representative sample was robust against issues of missing data and alternative explanations. Although most people who experience psychotic symptoms in adolescence will not develop schizophrenia later in life, they can indicate an increased risk for such disorders.

It is important to remember that there is evidence linking T. Gondii exposure in pregnancy to risk of miscarriage and stillbirth or health complications in offspring. We therefore recommend that pregnant women should continue to avoid handling soiled cat litter and other sources of T. Gondii infection (such as raw or undercooked meats or unwashed fruit and vegetables). That said, data from our study suggests that cat ownership during pregnancy or in early childhood does not pose a direct risk for later offspring psychotic symptoms.

PsyLife attends MQ Mental Health Science Meeting

By Jean Stafford

This month the PsyLife team attended the annual MQ Mental Health Science Meeting. Following on from MQ’s recent ‘We Swear’ campaign (right), the meeting was focused on the role of research in improving our understanding of the aetiology, treatment and prevention of mental health problems. The event was interdisciplinary with speakers and attendees including psychologists, neuroscientists, epidemiologists, clinicians, policy makers, and mental health service users united by a shared interest in mental health.

Several speakers presented findings about risk factors for mental illness. We heard from Professor Louise Arsenault and Dr Jean-Baptiste Pingault about the association between bullying victimisation and future mental health problems. Both talks were very interesting, and highlighted the need to focus on both the experiences of being bullied as well as predictors of bullying itself. We also heard from Professor Ezra Susser, one of psychiatry’s pre-eminent epidemiological thinkers, on his work examining the potential role of prenatal micronutrient deficiencies in neurodevelopmental disorders in offspring. Professor Susser highlighted the need for randomised controlled trials with long-term follow-ups, and/or the use of techniques such as Mendelian randomisation to further explore this topic.

Numerous talks focused on depression in young people. We heard about Professor Myra Weissman’s pioneering work showing the transmission of depression across three generations. Dr Frances Rice discussed antecedents of depression in at-risk adolescents and highlighted the importance of future studies examining the role of social support in this group.

Several talks demonstrated the role that neuroscience can play in improving our understanding of mental health problems and developing interventions. Dr Susanne Ahmari presented her work on the neural circuits underlying obsessive-compulsive disorder and how this could translate to future treatments. Professor Carmen Sandi discussed the impact of genes and environment on brain development, demonstrating that rodents exposed to early life stress showed structural and functional brain alterations and were hyper-aggressive later in life.

Professor Jonathan Mill discussed how epigenetic research could help to improve our understanding of mental illness, but also warned us not to believe all the media hype. To much amusement and shock from the audience, Professor Mill presented some of the media misrepresentations and pseudoscience surrounding epigenetics, including the advertisement of epigenetic spray and an epigenetic orthodontist!

It was interesting to hear several talks on psychosis from different perspectives. Professor Mary Cannon showed that children and adolescents who experience psychotic symptoms are at increased risk of developing future mental health problems, including schizophrenia, although these symptoms have a low positive predictive value and intervening early could therefore be problematic. Dr Graham Murray showed that, in people with chronic schizophrenia, the jumping to conclusions bias can largely be attributed to noisy decision making, whereas those with first-episode psychosis seem to genuinely jump to conclusions: perhaps due to perception of a high cost of gathering information.

Of great interest to our lab was Professor Andreas Meyer-Lindenberg’s talk on the relationship between the social environment and the brain. The talk focused on urbanicity and migrant status: two well-established environmental risk factors for psychosis identified in epidemiological research. Urban birth, urban living, and migrant status were found to be associated with structural and functional differences in brain regions implicated in social stress processing (such as the amygdala and the perigenual anterior cingulate cortex). These findings demonstrate how associations observed in epidemiological research can be further examined using psychology and neuroimaging techniques. However, participants in most of these studies were from the general population. It would be of great interest to see whether findings replicate in those with psychosis.

One further highlight of the conference was the thought-provoking panel discussion entitled ‘What good is a diagnosis?’. Megan Haste, a mental health blogger and service user, described the relief she felt when she was given a diagnosis, and Professor Susser pointed out that diagnoses can be helpful for constructing a personal narrative around mental health problems. However, he also discussed the challenges of using the same diagnoses across different cultures given cross-cultural variation in perceptions of mental illness. Professor Wessely highlighted the importance of diagnoses for providing appropriate treatments for patients, whereas Professor Wolpert suggested that it may be more beneficial to focus on a given patient’s symptoms and treatment needs, rather than categorisation.

The event also included a poster session where early career researchers presented work spanning a wide range of topics. I presented findings from a review of the incidence of very late-onset psychotic disorders, which I am carrying out as part of my PhD. This provided an opportunity to receive feedback on my work and to hear about current research on psychotic disorders, as well as other mental health problems, from across disciplines.

Overall, the meeting was a great opportunity to hear about some of the latest research targeting mental health problems in young people. The importance of addressing this issue is discussed from a clinician’s perspective by Dr Daria Monteforte, a psychiatrist from Verona currently on a research visit at PsyLife:

“As a medical doctor working in the field of mental health I really appreciated the choice of subject and the interventions discussed. It is now widely understood that psychiatric disorders often affect patients during youth, sometimes from childhood, and that only a small proportion of those in need are diagnosed and treated. A delay in intervention can undermine future opportunities for young people because of the destructive impact that mental health problems can have on many areas of life, including: building relationships, school, university, and career. More and more patients and their relatives are struggling with barriers to accessing services, diminishing resources in mental health services, and concerns about stigma. We work with many young adults who may have had a better prognosis if they had received the support they needed as children and adolescents. Research and discussion about mental health in young people helps us to understand the scale of the problem, what these particular patients need, and in which direction we should focus our efforts.”



Psychosis in the field

Click to visit the study webpage (external)New research led by Dr James Kirkbride from the PsyLife group, Division of Psychiatry, UCL, has shown how psychosis risk varies in more rural populations than previously studied in England. The findings, published last week in the American Journal of Psychiatry, reveal that rural regions see a significant rate of new cases of psychotic disorder in their populations. This is important because service planners need to ensure all people developing a first episode of psychosis can get quick, timely access to specialist mental health services such as Early Intervention teams.

The study was conducted in the East of England, in three counties – Cambridgeshire, Norfolk and Suffolk – and tried to identify all young people (aged between 16 and 35 years old) presenting to mental health services for the first time due to psychosis over a 3.5-year period. During this time, the study identified over 680 people who met diagnostic criteria for first episode psychosis living in this region, corresponding to an incidence rate of 34 new cases per 100,000 people per year.

The study also revealed that the risk of developing psychosis varied by important characteristics of both individuals and their environments. For example, although the study was conducted in a more rural setting than many previous epidemiological studies of this type , the authors found that the rate of disorder still varied according to how densely populated and deprived the environment was. Rates of psychotic disorder were relatively stable in most parts of the region, but rose in the most deprived and most densely populated regions. While this data is in keeping with earlier research , the study extends previous knowledge by suggesting that there may be a threshold effect in operation; below a certain level of population density or deprivation, there is no increase in risk, but once a certain level is reached, risk begins to shift. This research has not previously been possible in studies which have tended to be conducted in more urban environments.

The study also found variation by individual level factors. As previously established, men were almost twice as likely to experience a first episode of psychosis than women, particularly in their late teens and early twenties when risk is highest for both sexes. There was some evidence that rates were also raised in black and minority ethnic groups, but another paper from the same study will address this issue in more detail, and is currently under review.

You can read the full version of these findings here, or read about further overage of the findings from Psychiatric News and Medscape.


A Big Week for PsyLife

336437537It’s been quite a week for the PsyLife lab. First off, we’re delighted to say that one of our PhD students – Jen Dykxhoorn – successfully passed her PhD upgrade at UCL. While we weren’t surprised by this, it reflects a huge amount of work Jen has put in to her PhD on migration and psychosis in the first year. She’s currently writing up some results from her first study on this topic, which we’ll be submitting soon for publication. Jen has developed a great program of research as part of her PhD over the next 3 years, which we’re very excited about. Congrats Jen! Jen’s research is funded by Mental Health Research UK, with subsidiary funding from the Royal Society and Wellcome Trust.


We’re also delighted to announce that another PhD student – Jean Stafford – has joined the group to do a mixed epidemiological and psychological PhD on old age psychosis. She’ll be co-supervised by Dr James Kirkbride and Prof Rob Howard in the Division of Psychiatry. Jean spent 3-months with us last year as part of her MRC-sponsored first year PhD rotation scheme. Welcome along Jean!


We also welcome Dr Daria Monteforte from the University of Verona who joins us for a 6-month research visit. Daria is a psychiatrist specialising in psychosis. She’ll be gaining research experience during her visit with us. You can read more about what Daria’s up to here.


The PsyLife group would also like to pass on its congratulations to the MSc students who did student projects with us in 2015-16. All the projects we supervised addressed interesting epidemiological and mental health service issues with respect to psychosis, and we were impressed by the overall quality. Hopefully several of these will be submitted for peer-reviewed publication in the coming months. Dawid Gondek begins a PhD in longitudinal epidemiology of healthy ageing at the Institute of Education with Professor George Ploubidis, while Lucy Richardson has joined the REACH study with Professor Craig Morgan at the IoPPN as a Research Assistant. Mohammedi Abdolali continues his clinical work as a psychiatrist, armed with new insights from his MSc experiences. Finally, well done to Tom Steare, who continues his studies in mental health.


On a final note, a major publication from the PsyLife group has been published this week in the American Journal of Psychiatry. The paper investigates the epidemiology of first episode psychotic disorders, as seen through Early Intervention in Psychosis services in a rural population in England. Until now, very little epidemiological research has been conducted in EIP settings or in rural populations, and this study is an attempt to fill this gap. The paper has also been covered by Psychiatric News.

More Student Success in the PsyLife Lab

Bio picHere in the PsyLife lab we busy working away on lots of exciting hypotheses regarding how the social environment is linked to a greater risk of experiencing psychosis at a later time point.

Now, one of our first-year PhD students, Jen Dykxhoorn has achieved further recognition for her PhD which will focus on the role of social factors underlying the raised rates of psychotic disorders amongst migrants and their descendants. Jen has been awarded a prestigious UCL Overseas Research Scholarship in support of her PhD.

We’re delighted Jen’s research potential has been recognised at this early stage in her career, and we are excited about the progression she will make in her studies over the next three years. Jen’s PhD research is also supported by the Mental Health Research UK (MHRUK)’s John Grace QC Scholarship. Once again, congratulations to Jen, and stay tuned here for exciting findings from her and the rest of the group.

SIRS Travel Award for PsyLife PhD student!

SIRS meetingThis weekend will see the start of the eagerly awaited 5th Biennial Schizophrenia International Research Society (SIRS) conference, taking place in Florence, Italy. The conference will bring together international researchers from all areas of schizophrenia research to discuss the latest findings from the field. It promises to be an excellent meeting, and two members of the PsyLife group, Dr James Kirkrbide and PhD student Hannah Jongsma, will be presenting some of their findings.

We’re delighted to announce that Hannah has won a prestigious SIRS Young Investigator Travel Award to attend the conference. Congratulations! Hannah is in the second year of her PhD on the role of social and environmental risk factors in schizophrenia and other psychoses. To receive this international award is a major achievement and highlights Hannah’s dedication and growing expertise in the field. Hannah will be presenting her findings on cross-national variation in the incidence of psychotic disorders using data from the EU-GEI study. If you are at the conference, her talk will be on Tuesday 5th April, from 3.15pm, in the session on “Epidemiology: roles for environmental risk factors”. James will be presenting his work on age-at-migration in the same session. Do join us if you can.

See you in Florence!



Full list of PsyLife presentations at SIRS 2016

  • Hannah Jongsma: Incidence of psychotic disorders in England, France, Italy, the Netherlands, Spain and Brazil: Data from the EU-GEI study, Tues 5th April, 4.30pm-4.45pm: Symposium: Epidemiology: roles for environmental risk factors
  • James Kirkbride: Age-at-migration and risk of first episode psychosis in England: epidemiological evidence from the SEPEA study, Tues 5th April, 3.45-4pm, Symposium: Epidemiology: roles for environmental risk factors
  • James Kirkbride: (Poster S157): The Epidemiology of first episode psychosis in early intervention  in Psychosis services: findings from the SEPEA study, Poster session 1, Sunday 3rd April, 11am – 1pm.


Psychosis incidence in refugees to Sweden – cohort study of 1.3 million people

Working with colleagues at the Karolinska Institute we’ve shown that the risk of schizophrenia and other psychotic disorders is raised in refugees to Sweden by up to 3 times that of the Swedish-born population. This risk, was on average, 66% higher than the risk in other (non-refugee) migrants from the same regions of origin. These findings suggest that the additional psychosocial adversities faced by refugee groups may be important in the aetiology of psychosis. You can see a video abstract highlighting the work here, or read the full article published in the BMJ. This work was jointly conducted by our group at UCL and the Department of Public Health Sciences at the Karolinska Institute.

Changing childhood environments & behavioural outcomes in adolescence


Does the neighbourhood you live in during childhood lead to differences in behavioural outcomes in adolescence? This was the central question asked in some new research we have conducted in collaboration with Prof. Ian Colman and colleagues from the APEAL Lab, University of Ottawa, Canada.

Using Canadian data from the National Longitudinal Study of Children and Youth, on over 5000 children aged between 0-3 years in 1994-5, followed every two years until 12-15 years old, our research found that changes in the reported level of neighbourhood cohesion during childhood (as assessed by the primary caregiver, usually the mum or dad) led to changes in the risk of both pro-social and adverse behavioural outcomes later in adolescence.

Specifically, we found that children whose parents reported a decline in neighbourhood cohesion over time had a 67% greater risk of experiencing hyperactivity, while children whose neighbourhoods improved over time had lower risks of both hyperactivity and indirect aggression, compared with children who remained in stable cohesive neighbourhoods. Perceived improvements in neighbourhood cohesion during childhood were also associated with more pro-social behaviour in adolescence.

Our analyses tried to take into account as many alternative explanations for these findings as possible, including factors such as gender, family socioeconomic  status, neighbourhood deprivation, stressful life events, maternal depression & alcohol abuse. These factors did not explain the findings. Limitations of our study include the fact that we could not distinguish between children who moved to new neighbourhoods from children who remained in the same locality but whose neighbourhood environment changed over time.

Nevertheless, our research suggests that neighbourhood environments in childhood may shape future behavioural health outcomes and could support future interventions which aim to regenerate socially and economically deprived communities.

You can read the full article, published as a peer-reviewed FirstView paper in Psychological Medicine .

You can also read more about my group’s work on mental health and neighbourhoods .

3-year post-doc position in psychiatric epidemiology available in PsyLife group

PhD advertPosition now closed.

Post-doctoral Research Associate in Psychiatric Epidemiology

UCL Division: Division of Psychiatry

Specific Unit: PsyLife group

Grade: 7

Salary: £33,353 – £37,152 per annum (inclusive of London allowance)

Duration: Up to 3 years, subject to UCL probationary period

Application deadline: 17th May, 2015 

Application process: Via UCL main vacancies site Closed

Line Manager: Dr James Kirkbride (CI, PsyLife group)


Duties & Responsibilities

An enthusiastic, motivated individual is sought for this exciting early-career (postdoctoral) research opportunity in psychiatric epidemiology, available for a fixed period of 3 years in the PsyLife group, Division of Psychiatry at UCL. Our group’s interests include understanding the role of social and economic determinants on risk of psychotic disorders over the life course, with specific focus on societal or neighbourhood-level effects and risk amongst migrants and their descendants.

The post-holder will conduct research using two major epidemiological datasets (Swedish National Register data and the ALSPAC birth cohort) to investigate the social, spatial and life course epidemiology of psychotic disorders. There will be scope within the role for the post-holder to formulate their own research questions for analysis, concomitant with the interests of the PsyLife group. The post-holder will be expected to contribute to all major aspects of epidemiological research, including data preparation, study design, analysis of longitudinal and multilevel epidemiological data, critical interpretation of results, manuscript preparation and dissemination of research findings for peer-reviewed publication and translation to public, policy and clinical settings.

In addition to research duties, the post-holder will be expected to support junior colleagues in the PsyLife group and assist with some epidemiological teaching on our MScs in Clinical Mental Health and Mental Health Research.

The appointment is funded by the Wellcome Trust and the Royal Society.


Key Requirements

The successful applicant will have a PhD in epidemiology, biostatistics or a closely related discipline, as well as experience in conducting statistical analyses of large, epidemiological datasets, and in working with longitudinal or multilevel data. The post-holder will have demonstrable evidence of publication in peer-reviewed journals; candidates with a research profile approaching the national level are particularly desirable. The successful applicant will be experienced in aspects of database management and have a thorough working knowledge of at least one major statistical package (i.e. Stata, M-Plus, R, SAS or equivalent).



About the Division of Psychiatry and UCL

UCL is the leading UK university in terms of research power in the 2014 Research Excellence Framework, and is an international leader in biomedical science. Our Division is in the Faculty of Brain Sciences which has the highest level of ISI citations in Europe, is ranked second worldwide in neuroscience & behaviour and first nationally according to research power (REF 2014). We have an outstanding, multidisciplinary environment for research study, with internationally renowned experts in areas relevant to this proposal, including psychiatric epidemiology (Lewis, Kirkbride, King, Johnson, Osborn), health services research (Killaspy, Johnson), social psychiatry (Bebbington) and biostatistics (Lewis, Kirkbride). We have strong links with other Divisions, including Epidemiology & Public Health, Primary Care & Population Health, the Institute for Clinical Neuroscience and the newly-formed Institute of Clinical Trials and Methodology. This will place the student at the heart of our multi-disciplinary approach to psychiatric epidemiology. We have an excellent Divisional track record in postgraduate training with approximately 35 new taught MSc students and 3 new PhD students per year. Our on-time completion rates average above 90%, well above the Research Councils’ projected figure for PhD completion at any time (80%).


Page 1 of 212