Email Icon Facebook icon Twitter Icon GooglePlus Icon Contact

User Top Menu

ICOSR 2015—Social Cognition in Schizophrenia: Mechanisms, Course, and Treatment

25 May 2015

As part of our ongoing coverage of the 2015 International Congress on Schizophrenia Research (ICOSR), held March 29 through April 1 in Colorado Springs, 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 conference directors Carol Tamminga and Chuck Schulz, as well as meeting staff Cristan Tamminga and Dorothy Denton. For this report, we thank Anne-Kathrin Fett of the VU University Amsterdam, the Netherlands.

May 26, 2015. On the morning of Wednesday, April 1, 2015, the International Congress on Schizophrenia Research featured a symposium that integrated several different aspects of social cognitive functioning in schizophrenia, drawing together state-of-the-art findings from epidemiology, social neuroscience, and treatment research. Avraham Reichenberg of Mount Sinai Hospital in New York City chaired four presentations that addressed the following key questions:

1. What areas of functioning are impaired and when does this impairment start?


2. How are social cognitive deficits reflected in impairments during social interactions?


3. Is social cognition amenable to change?


4. What methods may be most successful?

Eva Velthorst, also of Mount Sinai Hospital, addressed the first question. She presented two studies on the trajectories of social functioning before and after the onset of psychotic illnesses. The first study included population-based data from more than 3,900 schizophrenia cases and their healthy siblings with an increased genetic illness risk. The results showed that in schizophrenia subjects, social engagement started to diverge from that of their unaffected siblings (who showed lower functioning than controls) 15 years prior to first hospitalization with a continuous downward trajectory. In contrast, impairments in functioning in structured environments and individual autonomy showed a decline that occurred closer (eight and three years, respectively) to first hospitalization. These findings stress the need for investigations that focus on different aspects of social outcomes in isolation in order to fully understand social impairment in schizophrenia. Velthorst's second study investigated the 20-year longitudinal trajectories of social engagement in a dataset that included more than 400 cases with schizophrenia, psychotic bipolar disorder, and psychotic major depressive disorder. Latent class analysis yielded four functional trajectories across psychotic disorders that had high temporal stability after illness onset, and that included classes with profound, severe, and moderate social impairments, and a functionally preserved class. Suggestive of a continuous representation of social functioning, each diagnosis was represented in three of four trajectory classes.

The presentation by myself, Anne-Kathrin Fett of the VU University Amsterdam, the Netherlands, dealt with the second question and brought the focus from the macro- to the micro-level of social functioning. My colleagues and I used an interactive trust game paradigm from game theory in subjects with early and chronic psychosis and healthy controls to investigate the mechanisms of disturbed social functioning during "online" social interactions. In line with our previous research, our new findings showed that psychosis was associated with reduced trust toward unknown others at the onset of social interactions regardless of illness phase. Subjects in the chronic illness phase did not learn to trust during repeated interactions with a benevolent, trustworthy other. Subjects with early psychosis, in contrast, were sensitive to others' positive social signals, as indicated by an increase in the levels of trust compared to those of controls as the interaction progressed. Interestingly, schizophrenia subjects and controls lowered their levels of trust similarly in response to betrayal by an unfair other. In other words, learning from negative social signals and the ability to adapt behavior in response seems to be intact in psychosis. These findings suggest that a loss of basic trust, paired with insensitivity to positive social signals, aggravates social dysfunction, particularly in the chronic illness stages, and that the early illness stages may present a window of opportunity for interventions that aim to keep the behavioral flexibility toward others intact.

Questions three and four were addressed in the second half of the symposium, which focused on new pharmacological and psychological treatments that seek to improve social cognition and functioning in schizophrenia.

Mark Weiser of the Sackler School of Medicine, Tel Aviv, Israel, presented newly unblinded results of an RCT on the effect of oxytocin as an adjunct treatment to social skills training and supportive therapy. Patients took part in social skills training or supportive therapy while using intranasal oxytocin or placebo three times daily for three weeks. The primary outcome measure was a sophisticated, structured assessment of social interactions. The quality of the social interactions, specifically focusing on gaze to the experimenter's face, vocalization (patient's vocal output, positive/negative tone, and fluent speech) and affect, body tone, movements, and other nonverbal signals, was assessed by raters who were blind to treatment status. The study hypothesis was that the combination of social skills training with oxytocin would demonstrate a distinct pro-social effect. However, the results showed no facilitating effect of oxytocin above placebo in either the social skills training or the support therapy group. Nonetheless, Weiser cautioned the audience against drawing quick conclusions based on these findings because, to date, the evidence from other research on the effects of oxytocin as a social (cognitive) enhancer has been mixed. He concluded that we still know little about how oxytocin actually works and that future research should address crucial questions such as the optimal dosage, the effects of single versus multiple administration, and effects on different groups such as first-episode and chronic schizophrenia patients.

Completing the symposium, David Roberts of the University of Texas, San Antonio, presented pilot data on a new social cognitive de-biasing intervention that combined iPad training with phone support. The intervention aimed to activate biases by increasing self-relevant processing and then to increase awareness of and control over biases by teaching participants an easy heuristic of three characters to recognize internal (blaming self), external (blaming others), and situational (blaming bad luck) attribution styles. Outcome was assessed in terms of feasibility, social cognitive performance, social engagement, and measures of attribution bias. Overall, the results showed that in-home iPad-based training is feasible in schizophrenia. A first study assessed the use of complex self-relevant (second-person) social stimuli to achieve a reduction in attribution bias and compared these to the use of simpler third-person stimuli. Unexpectedly, the training with third-person stimuli yielded somewhat better effects in terms of social engagement. Furthermore, there was an increase in hostile attributions associated with the use of self-referential stimuli.

A second study compared the training with self-relevant stimuli and phone support to a control group that played a computer game. The only significant difference in outcomes was that those in the training group showed an increase in hostile attributions. Roberts suggested that utilizing self-relevant stimuli might have been too complicated, or it could have activated biases as intended but then failed to neutralize them. While the training with third-person stimuli yielded tentative improvements in social engagement, he concluded that more work would be needed to strengthen the intervention's ability to increase patients' awareness of their biases and to dampen the effect of these biases on social behavior.

The discussant, Michael Green of the University of California, Los Angeles, placed key findings into a bigger picture. First, outcomes in social functioning are different from other types of community outcomes such as vocational functioning or independent living. Community functioning is often used as an umbrella term for a conglomerate of social outcomes in schizophrenia, but may fail to reflect the complexity of outcomes accurately. For instance, different outcome domains also seem to have different predictors. Vocational functioning and independent living are better predicted from cognitive test performance than social functioning. This, according to Green, highlights the importance of fine-grained investigations that take different outcomes into account. Second, contrary to popular assumptions, social reward learning does not seem generally impaired in patients with psychosis. Most crucially, in chronic patients, learning appears intact when it comes to negative social cues, but not positive ones. This asymmetry in learning from social feedback is vital for the understanding of social dysfunction and in line with studies on reward learning in nonsocial conditions, which showed better learning in patients in response to negative than positive feedback.

Third, Green remarked that research on oxytocin as an agent to alleviate social cognitive problems is still in its infancy. Research has yielded contrasting findings, which is not surprising given the variation in utilized study parameters such as dosage and administration patterns or outcome measures. For instance, the effects of oxytocin on social competence may differ from those on social cognitive task performance. In addition, very little is known about target engagement with oxytocin. This makes studies on oxytocin very hard to design. Thus, despite the negative findings in the reported trial, these are still the early days, and further work needs to be done. Finally, Green stated that a bias or belief might be viewed as a possible consequence of social cognitive problems rather than a fundamental cognitive deficit itself. So far, studies on bias modification have not been very successful. Possibly, the results could be improved if lower-level cognitive problems were targeted at the same time.

In sum, the presented findings showed long-standing impairments in social functioning in psychotic disorders that for some domains start decades before illness onset and suggest that social problems might be driven by the insensitivity to positive social reward. Unfortunately, the presented treatment interventions did not yield the much-hoped-for improvements. This stresses the need for more research that systematically investigates interventions that aim to ameliorate social (cognitive) dysfunction.—Anne-Kathrin Fett.