way emotion modulates cognition in recent onset schizophrenia patients (ROS). After developing a new task to measure the interaction between emotional expressions and sustained attention (chapter 2), we investi- gated whether this interaction is disturbed in recent onset schizophrenia (chapter 3). Next, we investigated the interaction between emotion and attention in recent onset schizophrenia by using the emotional Stroop task (chapter 4).
We reviewed the literature to see how methodological differences influence results when studying emotional memory modulation in schizo- phrenia (chapter 5). Using these findings we investigated if disturbances in emotional memory modulation are present in recent onset schizophre- nia patients and whether these disturbances are distinguishable between short- and long-term visual and verbal memory (chapter 6). Finally, we investigated social cognition in recent onset schizophrenia (chapter 7).
The influence of emotion on attention
Although there is a growing body of literature demonstrating both the beneficial (Anderson, 2005; Ohman, Lundqvist, & Esteves, 2001; Phelps, Ling, & Carrasco, 2006) and the interfering (Vuilleumier & Brosch, 2009) effect of emotion on performance in attention tasks, only a few studies investigated the influence of emotional expression on sustained attention. We designed a novel task in which positive, negative and neu- tral facial expressions were used as task irrelevant cues in a Emotional Continuous Performance Task (E-CPT) to study the influence of negative and positive emotion on sustained attention. First we compared the in- trinsic test properties of the E-CPT with those of a classic letter CPT (AX- CPT) and second we investigated the influence of emotional expressions on sustained attention. The results are described in chapter 2. Although accuracy was lower and reaction time (RT) was longer in the E-CPT com- pared to the AX-CPT, this was most likely due to using stimuli that were harder to discriminate within sets and did not change the intrinsic test properties. In our E-CPT we found differential effects of emotion on reacti- on time and accuracy; faces with positive and negative emotion improved reaction time for the subsequent probe faces in comparison to neutral fa- ces. This could be a result of affective priming (Palermo & Rhodes, 2007) which means that seeing an emotional face causes the subsequent neu- tral probe face to be perceived as the same emotion and therefore also receiving enhanced processing (Palermo & Rhodes, 2007). This effect is generally considered to be brief (200-500ms (Chun & Potter, 1995)), but affective priming can be prolonged beyond 500ms in conditions of low at- tentional load (Vuilleumier & Huang, 2009), such as in our study. Holmes, Bradley, Kragh Nielsen, and Mogg (2009) used a visual-probe task with angry, happy and neutral facial expressions as cues and non-emotional icon as probes (arrows) to investigate investigated the temporal course of attentional biases for facial expressions in HC. In their task, a neutral and an emotional cue face were shown side-by-side followed by an arrow
8
at the location of either the emotional or the neutral face. Participants were required to indicate which way the arrow was pointing. Just like in our study Holmes et al. (2009) found faster RTs when the probes ap- peared at the location of the emotional cue faces compared to when the probes appeared at the location of the neutral faces. The authors conclu- ded that this is in line with an attentional bias for emotional compared to neutral faces. Because they also used electrophysiological measurements they could demonstrate that the attentional orienting to emotional faces began earlier for negative than positive faces (Holmes et al., 2009).
In our study the effect of facial emotional expressions on accuracy was specific to negative emotion; accuracy decreased after negative, but not after positive or neutral cue faces. Since negative stimuli capture more attention than positive stimuli (Eastwood, Smilek, & Merikle, 2003), negative cue faces might interfere with the processing of subsequent pro- bes by placing a heavier demand on attentional resources (Yiend, 2010), resulting in lower accuracy. This reallocation of resources could also ex- plain better visual short-term performance for angry compared to neutral and positive faces, even when facial expressions are a task irrelevant stimulus property (Jackson, Wolf, Johnston, Raymond, & Linden, 2008). In our study, emotion effects on probe perception were only present at the short (500ms) inter-stimulus interval and not in the long (5000ms) inter-stimulus interval. This is in line with the study by Ciesiellski et al. (2010) who studied the temporal course of attention-emotion interactions and found that emotions impede accuracy more, when the time between the emotional distracter and the target stimulus is relatively short. This means that for negative faces there is a speed-accuracy trade off; they induce a mode of processing focused on reacting fast but not especially accurate, which is in line with evolutionary explanations for an emotion- attention interaction. From an evolutionary perspective it is useful to in- crease processing only after a signal that something important is about to happen when sustained attention gets less efficient due to fatigue effects. Positive stimuli however speed up processing, but without interfering with efficiency; they induce a mode of processing focusing less on speed and more on evaluation of the surrounding stimuli, as is predicted by Frede- rickson’s (Fredrickson, 1998) broadening hypothesis.
Patients suffering from schizophrenia have impairments on several emotional domains. Many of them report anhedonia (Burbridge & Barch, 2007) and they are often impaired in maintaining emotional, especial- ly positive, experience over longer periods of time (Kring & Caponigro, 2010). They have difficulties in using facial cues for social judgment (Mar- wick & Hall, 2008) and often show reduced facial expressions themselves (Mandal, Pandey, & Prasad, 1998). They also suffer from difficulties in processing emotional expressions of human faces (e.g. Kohler, Walker, Martin, Healey, & Moberg, 2009; Morris, Weickert, & Loughland, 2009) which is independent from a generalized impairment in face perception (Goghari, Macdonald, & Sponheim, 2010). Nevertheless, at the moment of exposure to emotional stimuli the valence ratings and physiological