Email Icon Facebook icon Twitter Icon GooglePlus Icon Contact

User Top Menu

ICOSR 2017: Risk Factors and Vulnerable Minds

27 Apr 2017

As part of our ongoing coverage of the 2017 International Congress on Schizophrenia Research (ICOSR), held March 25-28 in San Diego, we bring you session summaries from some of the participants in the Young Investigator program. We are, as always, grateful for the gracious assistance of YI program directors Laura Rowland and Scott Sponheim, as well as Michelle Tidwell of the ICOSR staff. For this report, we thank Alexis Cullen of the Institute of Psychiatry, Psychology and Neuroscience at King's College London.


Over the past two decades, schizophrenia researchers have increasingly focused on the putatively prodromal phase of psychosis, as identifying individuals in this phase may provide the best opportunity for preventative intervention. In conjunction with these efforts, researchers have continued to search for risk factors for schizophrenia and elucidate the mechanisms by which environmental and biological risk factors might converge to give rise to psychosis. At the 2017 ICOSR meeting, a session dedicated to "Prodrome and Risk" showcased advancements in these two research streams.

Lotta Kinnunen of the University of Oulu in Finland presented data from the Northern Finland Birth Cohort 1986 Study (N = 6,682), an impressive longitudinal research program comprising healthcare records, questionnaire data, and biological samples. Analyses were conducted to examine the relationship between parental somatic illness and prodromal symptoms of psychosis (assessed via the PROD-screen questionnaire) at age 16 years. Of the 19 parental disorders examined, only musculoskeletal disorders were associated with the presence of prodromal symptoms in offspring. Results were largely unchanged after adjusting for potential confounders, including parental psychiatric disorders, and the relationship was stronger in male offspring. Further research is needed to determine whether this positive association might be explained by parental psychiatric symptoms (i.e., symptoms that do not meet the threshold for a  disorder), discontinuity of parenting (i.e., due to hospitalization), or shared genetic factors.       

The second presentation in this session, delivered by Charles Davidson of Yale University, examined trajectories of social cognitive function among individuals at clinical high risk (CHR) for psychosis recruited to the North American Prodrome Longitudinal Study (NAPLS). Analyses conducted on cross-sectional data indicated that emotion perception, social perception, and theory of mind improved with advancing age in both CHR individuals and healthy controls. Whilst trajectories for emotion and social perception abilities did not differ by group, individuals at CHR had a lower slope (indicating a blunted increase relative to healthy controls) for theory of mind ability. Interestingly, this smaller slope was only true for CHR individuals who were antipsychotic and antidepressant naïve, suggesting that pharmacological treatment may help to improve theory of mind in those at risk for psychosis. There was no evidence to suggest that duration of psychosis had an effect of social cognition. Further research, utilizing repeated measurements of social cognition over time, is needed to fully understand these cross-sectional associations.

Utilizing data from the Australian Personal Assessment and Crisis Evaluation (PACE 400) study, a longitudinal study of individuals at CHR, Scott Clarke of the University of Adelaide, Australia, presented data on the prediction of transition to psychosis. Over the 13-year follow-up, just over one quarter of the sample developed a psychotic disorder; global functioning, duration of symptoms prior to presentation, quality of life scale, disordered thought content, conceptual disorganization, performance IQ, and full-scale IQ were found to significantly predict transition. Bayesian modeling was used to combine positive and negative likelihood ratios for these variables, and the resulting model predicted transition to psychosis with sensitivity and specificity values of 50-64 percent and 75-92 percent, respectively (positive predictive value: 50.7 percent; negative predictive value: 86.4 percent). Predictive accuracy was greatly improved for the Ultra-High Risk of psychosis (UHR)―Brief Limited Intermittent Psychotic Symptoms (BLIPS) group compared to the attenuated symptom and vulnerability groups. These findings demonstrate that clinical variables can be used to improve prediction of transition to psychosis.

Rachel Loewy of the University of California, San Francisco, next focused on exposure to trauma and how this might be associated with biological responses to psychosocial stressors in a large sample of CHR individuals (N = 102) and healthy controls (N = 64). Exposure to narrowly defined trauma (i.e., events that would be considered sufficient to precipitate post-traumatic stress disorder) was more common among individuals at CHR relative to controls (61 vs. 30 percent). Furthermore, transition to psychosis at 12 months was predicted by exposure to trauma at baseline, with larger effect sizes observed for traumatic events occurring prior to 12 years of age. Salivary cortisol samples acquired at awakening (considered to be a minor stressor) and during the Trier Social Stressor Test (TSST) were used to assess changes in hypothalamic-pituitary-adrenal (HPA) axis function in response to psychosocial stress. Compared to healthy controls, CHR individuals showed a statistical trend for elevated cortisol responses upon awakening and in response to the TSST. An interaction was observed with exposure to trauma whereby cortisol levels at awakening were significantly elevated among CHR individuals with a history of trauma relative to those without a history of trauma. Similar to findings obtained in adult studies, these data indicate that early trauma exposure can predispose the HPA axis to respond maladaptively to future stressors.

Moving to an earlier stage of psychosis risk, Colm Healy of the Royal College of Surgeons in Ireland presented data on the presence of psychotic-like experiences (PLEs; i.e., subclinical psychotic symptoms) in a longitudinal cohort of adolescents. Factor analysis conducted on PLE interview data obtained from 212 adolescents at baseline indicated the presence of two distinct factors: hallucination-like experiences (HLEs; e.g., seeing and hearing things that other people do not) and delusion-like experiences (DLEs; e.g., feeling that people can read one’s mind). At follow-up (approximately four years later), HLEs but not DLEs were found to predict total levels of psychopathology (as assessed via the Youth Self-Report) and internalization of problems. In contrast, global functioning at follow-up was more strongly associated with DLEs at baseline, and only DLEs (not HLEs) predicted the most severe functional impairments. These findings highlight the divergent contributions of hallucinatory and delusional experiences in early life for clinical and functional outcomes.

The next presentation of this session, by Marta Di Forti of the Institute of Psychiatry, Psychology & Neuroscience at King's College London focused on the relationship between genetic risk (as determined via polygenic risk scores) and exposure to cannabis use in a large sample of first-episode psychosis patients and healthy controls (N = 2,300). Polygenic risk scores across the total sample were categorized into quartiles; logistic regression analyses indicated that the odds of having a diagnosis of psychosis were seven times higher among individuals with scores in the highest quartile relative to the lowest quartile. In contrast to hypotheses, regular users of high-potency cannabis did not differ on polygenic risk score from occasional users or individuals who had never used cannabis. Moreover, there was no evidence of an interaction between cannabis use and polygenic risk score when predicting case status (first-episode vs. healthy controls). Whilst these results suggest that genetic predisposition to psychosis does not influence tendency to use cannabis, further research is needed to determine whether specific gene profiles (as opposed to global polygenic risk scores) might be associated with cannabis use.

The final presentation of the session, delivered by Mohini Ranganathan of Yale University, focused on the role of the endocannabinoid system in those at increased risk for psychosis as conferred by a family history of the disorder. Using a range of experimental techniques, researchers examined functioning of the endocannabinoid system in a small sample of individuals with a family history of psychosis and those without a family history. In the first study, high-resolution PET techniques using [11C]OMAR demonstrated that individuals with a family history of psychosis had lower cannabinoid 1 receptor (CB1R) availability compared to those without a family history. A second study, investigating the effect of intravenous THC administration using a randomized, placebo-controlled, double-blind design, indicated that individuals with a family history of psychosis demonstrated a higher increase in positive symptoms (as measured by the PANSS) in response to THC compared to those without a family history. Finally, those with a family history of psychosis also showed increased THC-induced cortical noise (also known as neural noise). These interesting findings require replication in a larger sample.

The session closed with an interesting discussion led by Daniel Mathalon of Stanford University who acted as chair. Members of the audience, including Robin Murray and Mary Cannon, commented on the importance of the findings and suggested directions for future research.