Email Icon Facebook icon Twitter Icon GooglePlus Icon Contact

User Top Menu

Social cognition in schizophrenia.

Green MF, Horan WP, Lee J
Nat Rev Neurosci. 2015 Oct; 16(10):620-31. PMID: 26373471. Pubmed


Submitted by Anouk van der Weiden on

This comment was co-written by Anouk van der Weiden, Merel Prikken, and Neeltje E.M. van Haren.

Why schizophrenia patients show social cognition abnormalities

Green and colleagues offer a great overview of the current state of research on social cognition in schizophrenia, covering perception of social cues, experience sharing, mentalizing, and experiencing and regulating emotion, from a mostly neural perspective. The review clearly indicates in what domains schizophrenia patients show abnormalities compared with healthy controls, which is important for guiding future research and treatment. However, the review article does not cover why and under which circumstances patients show abnormalities in these domains. Hence, we would like to take this opportunity to highlight a few well-studied processes that are involved in these domains as well as some modulatory factors, taking a more social cognitive perspective.

1. Mirror neuron functioning―Reduced anticipation or reduced distinction?

With regard to motor resonance, or mirror neuron functioning, the authors conclude that the evidence is mixed, possibly because of diverse scientific approaches. However, if you take into account the different aspects of mirror neuron functioning, i.e., anticipation and distinction, this may explain part of the inconsistency (see also van der Weiden et al., 2015). According to current perspectives, the mirror neuron network is not involved in simple motor resonance, but rather plays an essential role in the anticipation, understanding, and integration of one's own and other people's actions (Bonini et al., 2013; Ocampo and Kritikos, 2011). This can be assessed by measuring activation in the mirror neuron network during action observation. In addition, as the mirror neuron network is typically less active during action observation compared with action performance, it is also involved in distinguishing between one's own and others' actions (Mukamel et al., 2010). Such self-other distinction can be assessed by comparing activation in the mirror neuron network for observed versus self-produced actions. Considering these different measures, the empirical evidence quite consistently suggests that schizophrenia patients show abnormalities in both anticipation and distinction. That is, patients tend to show reduced activity in the mirror neuron network compared with healthy controls when observing other people's actions (i.e., reduced anticipation; Enticott et al., 2008; Mitra et al., 2014; Schurmann et al., 2007; Thakkar et al., 2014). Furthermore, they show reduced distinction between one's own and others' actions (i.e., undifferentiated activation, e.g., due to increased activation in areas that are normally inhibited; Jardri et al., 2011; McCormick et al., 2012; Thakkar et al., 2014).

2. Mentalizing―Simulation or interpretation?

As the authors conclude, the evidence on patients' mentalizing ability is inconsistent. We propose that these inconsistent findings may be explained by taking into account the different processes involved in mentalizing (see also van der Weiden et al., 2015). First, as mentalizing depends on motor simulation (cf. simulation theory) as well as cognitive belief processes (cf. theory theory) (Mahy et al., 2014), schizophrenia patients' abnormalities may come from different sources. The relative contribution of these processes depends on individual and contextual differences, e.g., in IQ and working memory load (Brune, 2003; Pousa et al., 2008), or processing speed (e.g., slowed emotion processing; Brennan et al., 2014). Because patients are capable of motor simulation (although to a lesser extent than controls), they perform relatively well in attributing basic mental states (Brune, 2003; Pousa et al., 2008; Shamay-Tsoory et al., 2007). Especially when situations get more complicated or demanding (and motor simulation becomes less informative), schizophrenia patients have difficulty understanding other people's mental states (Abu-Akel and Shamay-Tsoory, 2013; Corcoran et al., 1995; de Achaval et al., 2010). Second, as patients show abnormalities in motor simulation (i.e., mirror neuron functioning) in terms of anticipation (or understanding) as well as distinction, they may show associated mentalizing difficulties. Indeed, impairments in patients' mentalizing ability is not only apparent in misperception of other people's intentions or emotions (i.e., reduced anticipation or understanding; Derntl et al., 2009; Green et al., 2012; Haker and Rossler, 2009; Kring and Elis, 2013), impairments are also evident in patients' increased tendency to take over the (sometimes misperceived) intentions and emotions of others (i.e., reduced distinction, as in emotional contagion; Corbera et al., 2013; Decety and Lamm, 2011; McCormick et al., 2012; Montag et al., 2007; Ruby and Decety, 2004; Smith et al., 2015). Consequently, it is also important to take into account the role of anticipation and distinction in affective sharing, as schizophrenia patients may be 'sharing' misperceived (i.e., and thus not identical) emotions.

3. Empathy―Reduced ability or reduced motivation?

Empathy is often suggested to be impaired in schizophrenia, and the conclusion of Green and colleagues is no different. Surprisingly however, it remains unclear why patients show less empathy than healthy controls. First, empathy impairments may result from both motor and cognitive processes, and may involve anticipation versus distinction problems. That is, patients' abnormalities in empathy may be due to an inability to take the social context into account, or to abnormal cognitive theories or beliefs about the intentions or emotions of others. Second, it may be that patients' empathic ability is not impaired at all, but that their reduced empathic responses result from a decreased affiliative motivation. Depending on which of these processes are most responsible for patients' reduced empathic responses, fairly different interventions should be used when aiming to improve social functioning. Hence, it is important to take these factors into account and distinguish between them in our future research.

4. Emotion experience―Largely intact or significantly impaired?

The authors conclude that emotion experience in schizophrenia is largely intact, even though patients experience more negative affect in response to neutral and pleasant stimuli. Of course, such abnormalities in emotion experience may result from impaired emotion perception rather than an inability to experience emotion. However, patients' abnormalities in emotion experience cannot always be explained by impaired emotion perception. Converging on this notion, an extensive review on emotion perception, experience, and expression also provides a more nuanced picture (Kring and Elis, 2013). For example, when looking back on their own emotional experiences of the past week, schizophrenia patients' experiences are relatively weak compared with those of healthy controls, and this is not related to recency of the emotional experience (Ben-Zeev et al., 2012). We would like to stress that such group differences are not to be taken lightly, as they may have profound effects on patients' quality of life and social functioning.

5. Brain-behavior dissociations―Neural inefficiency or neural abnormality?

Finally, the authors discuss several studies where patients show discrepancies in neural versus behavioral performance. For example, patients show increased activation in the STG and PFC when inferring intentions and emotions of others, while their behavioral performance does not differ from controls. To account for such brain-behavior discrepancies, the authors offer a simple explanation, i.e., neural inefficiency (or overcompensation). This is not an uncommon explanation in fMRI research. However, it is a purely descriptive explanation that downplays or even discards the possibility that there are abnormalities in neural processing and behavior that were not measured. We would like to emphasize that although certain differences between groups in neural processing (e.g., hyper/hypo activation in patients) are not manifested in measured behavior, they may nevertheless be related to abnormal behavior or experiences that were not measured in the task at hand. We must go beyond simple descriptive explanations and develop testable paradigms to understand under which circumstances neural and/or behavioral performance is abnormal. Specifically, future research should administer a broader range of measures and manipulations in order to shed light on the extent to which behavior is normal under conditions of neural hyper/hypoactivation: How, when, and why are certain brain regions hyperactivated? How long lasts this hyperactivation? And how does it affect activity in other parts of the brain (e.g., that may be differently involved in social functioning)? For example, future research could assess participants' experiences during task performance. Also, by manipulating task demands we may establish when patients can no longer keep up with healthy controls because of their neural inefficiency. In doing so, it is important to take into account when and how neural and behavioral performance are measured. For example, it is not surprising to find abnormal behavior (e.g., more negative emotional experience in response to neutral and pleasant stimuli) but typical neural activation patterns, if neural and behavioral performance are assessed in different ways and at different moments in time (e.g., when participants are asked to reflect upon their emotional experience after [not during] the emotion perception fMRI-task; see Kring and Elis, 2013, for a similar argument).


Submitted by Deanna M. Barch on

Green and colleagues have provided an outstanding review of the extent literature on intact and impaired components of social cognition among individuals with schizophrenia. Perhaps even more importantly, they also highlight many of the critical "next steps" in this domain of research. These key next steps include examining these constructs transdiagnostically and examining the ways in which functional or structural connectivity between brain regions that play a role in social cognition contribute to intact versus impaired function.

Another key question that Green and colleagues allude to is the importance of understanding the degree to which impairments in social cognition and related neural systems are unique to the domain of social cognition, versus sharing mechanisms and processes with other known impairments in non-social cognitive function in schizophrenia. For example, in previous important work, Green and colleagues, as well as others, have shown that social cognition is an important mediator between deficits in "cold" cognition and functional outcome in schizophrenia. Further, it is also the case that social cognition accounts for variance in life function that is not accounted for by non-social cognition ("cold" cognition). Further, in this current review, Green and colleagues highlight the fact that many of the domains of social cognition impairment in schizophrenia are those that are reflective and which require effortful processing, a finding that is strongly reminiscent of the nature of impairments in "cold cognition" in schizophrenia, which are often strongest for those domains involving cognitive control or effortful processing. In addition, for some of these reflective components of social cognition, the literature on neural impairments in schizophrenia also points to regions important for cognitive control more generally, such as the dorsolateral prefrontal cortex. However, Green and colleagues also review literature demonstrating alterations in neural processing in regions that are not commonly reported in studies of non-social cognition in schizophrenia, such as the ventral medial prefrontal cortex. Thus, a critical pathway for future research will be to identify the psychological and neural mechanisms that make contributions to impairments in both social and non-social cognition, as well as the mechanisms that may make unique contribution to difficulties in one domain versus another.

This research can also incorporate examining of the role of connectivity abnormalities, as it is possible that some of the distinctions between social and non-social cognition arise from the differential patterns of connectivity that arise during the processing of social versus non-social information. Such an approach will require some adjustment to the research programs of many investigators, as there has been a tendency for particular labs to focus on social versus non-social cognition, but not necessarily both. Incorporating simultaneous examination of both in future studies will help to push forward this line of research and to identify core common and unique mechanisms that may be ripe targets for intervention, at either a psychological or neurobiological level of analysis.