Dysfunctional coping with stress in psychosis. An investigation with the Maladaptive and Adaptive Coping Styles (MAX) questionnaire

Abstract

Objective

Psychotic episodes have long been conceptualized as inevitable incidents triggered by endogenous biological impairments. It is now well-accepted that the ability of an individual to deal with social and environmental challenges plays an important role in regard to whether or not a vulnerability to psychosis translates into symptoms. For the present study, we examined symptomatic correlates of dysfunctional coping in psychosis and aimed to elucidate a profile of coping strategies that distinguishes patients with schizophrenia from those with depression.

Method

The newly devised Maladaptive and Adaptive Coping Styles Scale (MAX) was administered to 75 individuals with psychosis, 100 individuals with depression and 1100 nonclinical controls.

Results

Schizophrenia patients showed compromised coping abilities relative to nonclinical controls, particularly a lack of engaging in adaptive coping. Depression was more closely tied to dysfunctional coping than were positive symptoms as indicated by group comparisons and correlational analyses. Correlations between positive symptoms, particularly paranoid symptoms, and avoidance and suppression remained significant when depression was controlled for.

Conclusions

Although maladaptive and adaptive coping are unlikely to represent proximal mechanisms for the pathogenesis of positive symptoms, fostering coping skills may reduce positive symptoms via the improvement of depressive symptoms, which are increasingly regarded as risk factors for core psychotic symptoms. Furthermore, the reduction of avoidance and suppression may directly improve positive symptoms.

Introduction

The term schizophrenia was initially coined as a plural diagnostic entity (i.e., the schizophrenias) highlighting the heterogeneous phenomenology of the disorder (Bleuler, 1950). Core features encompass positive symptoms as well as negative, disorganized and affective symptoms (American Psychiatric Association, 2013). In view of the stigma associated with the diagnosis, which is often confused with split personality and equated with erratic and violent behavior (McNally, 2010, Schlier and Lincoln, 2013), clinicians have increasingly begun to abandon the term schizophrenia and to replace it with the yet unofficial label (non-affective) psychosis. Under the influence of the vulnerability-stress model, Japan even went a step further and renamed schizophrenia to integration disorder (“Togo Shitcho Sho”; Sato, 2006).

The vulnerability-stress model (Nuechterlein and Dawson, 1984, Zubin and Spring, 1977) posits that an acute episode only occurs if the level of stress exceeds the amount an individual can bear depending on his or her liability to psychosis. Despite the heuristic value of this account, the exact mechanisms that translate stress into psychosis have not been fully unraveled and research is beginning to focus more on predisposing factors that may act as moderators or mediators between stressors and symptoms (Lincoln et al., 2010, Lincoln et al., 2009, Moritz et al., 2015, Moritz et al., 2011).

Coping/emotion regulation is increasingly examined in psychosis. Coping is aimed at regulating emotional experiences by changing one's response to a stressful event (emotional coping/emotion regulation) or by changing the stressful situation itself (Compas et al., in press). Research traditionally distinguishes adaptive versus maladaptive coping (Aldao et al., 2010, Moritz et al., 2016a), whereby the specific context determines whether a style is functional or dysfunctional (Aldao, 2013). Coping has been most extensively researched in depression and anxiety and studies suggest that mood disorders are strongly associated with maladaptive forms of coping (Moritz et al., 2016), for example rumination (Olatunji et al., 2013). Whether or not a coping pattern is specific for a certain disorder is subject of an ongoing debate (Aldao and Nolen-Hoeksema, 2010, Aldao et al., 2010) and the question is complicated by the fact that most studies only assessed single disorders (Aldao, 2013). Complicating things further, we must take into account that coping styles are likely to be related to different symptoms for different reasons. Withdrawal/avoidance may foster depression and anxiety because of the feeling of loneliness but may also induce paranoia (Freeman et al., 2005), for example because correcting social feedback is lacking and correcting experiences are not made (Moutoussis et al., 2007).

Research into coping and emotion regulation in schizophrenia shows that patients engage in maladaptive coping, for example, excessive avoidance and safety behavior (Freeman et al., 2007) as well as emotion and thought suppression (Kimhy et al., 2012, Livingstone et al., 2009, van der Meer et al., 2009). Lack of awareness and tolerance of emotions is also prominent (Kimhy et al., 2012, Lincoln et al., 2015a, Lincoln et al., 2015b, van der Meer et al., 2009).

The urge to control (or suppress) one's (negatively valenced) thoughts or emotions has been linked to positive symptoms and particularly hallucinations (de Leede-Smith and Barkus, 2013, Jones and Fernyhough, 2006, Moritz et al., 2010, Morrison and Wells, 2003). This might be explained by the well-documented observation that suppression paradoxically enhances negative thoughts (Wenzlaff and Wegner, 2000). While in depressed patients this may increase depressive symptoms and helplessness, it could create a “tug war” with voices in those with psychosis (“once I try to resist the voices, the voices strike back and become louder”).

A recent review and meta-analysis (O'Driscoll et al., 2014) on coping in schizophrenia found that attentional deployment (e.g., rumination, worry), dissociation and alexithymia were positively associated with schizophrenia, whereas the ability to manage emotions was negatively associated with schizophrenia at a large effect size. Lincoln et al., 2015a, Lincoln et al., 2015b found that participants with psychosis reported a stronger increase in self-reported stress in response to a stressor than healthy controls which was in turn predicted by a reduced awareness of and tolerance for distressing emotions. Another study (Westermann et al., 2014) suggests that researchers should not only assess whether or not a coping strategy (e.g., reappraisal) is used but also if it is adopted successfully (see also Moritz et al., 2016); nonclinical delusion-prone individuals were found to be less successful in applying reappraisal (Westermann et al., 2014), which was, however, not confirmed in patients with manifest psychosis (Grezellschak et al., 2015).

The inferences that can be drawn from the available body of research are promising but limited by a number of factors. Firstly, the specificity of coping dysfunctions in psychosis is not fully established as individuals with psychosis were often tested against nonclinical (van der Meer et al., 2009) but not against clinical controls (e.g., depressive patients) (however see Kimhy et al., 2012, Lincoln et al., 2015a, Lincoln et al., 2015b) and the effects of general psychopathology, particularly depression, were not always considered (O'Driscoll et al., 2014). Secondly, results are inconsistent across participants with varying levels of liability to psychosis. As shown, nonclinical controls scoring high on psychosis proneness scales may display different responses and coping patterns than individuals who fulfill diagnostic criteria for psychosis requiring treatment (de Leede-Smith and Barkus, 2013). Of concern, the latter aspect, treatment status, may influence coping in opposing ways. Hospitalization may both promote certain dysfunctional coping strategies (e.g., passivity due to restricted and shielded environment) versus foster engagement in functional strategies (e.g., because of psychoeducational groups teaching patients new skills) thereby reducing the stability of findings. Thirdly, while research confirms that coping may play a role in the formation of paranoia (Westermann and Lincoln, 2011, Westermann et al., 2013, Westermann et al., 2012), the specificity of these findings is unclear.

We recently constructed a brief coping scale termed Maladaptive and Adaptive Coping Questionnaire (MAX; Moritz et al., 2016) which encompasses three subscales (maladaptive coping, adaptive coping, avoidance) covering a range of traditional coping styles (e.g., problem-solving, rumination) as well as styles of emotion regulation that have recently received more attention (e.g., acceptance). We developed a new scale rather than combining established ones as existing scales often tap single constructs and measures vary according to wording, response options, and also time frame (one week, lifetime etc.) which impacts on the respective reliabilities and validities. For this reason, correlations with, for example, specific psychopathological scales will mirror both conceptual but also methodological differences.

The MAX was recently validated in a large population sample (N = 2200) and found to yield good psychometric properties. The adaptive (α = 0.87) and maladaptive (α = 0.85) coping subscales showed high internal consistency; the avoidance subscale was somewhat less consistent (α = 0.65). The six-months test-retest reliability was especially high for maladaptive coping (maladaptive coping: r = 0.75, p < 0.001; adaptive coping: r = 0.61, p < 0.001; and avoidance: r < 0.59, p < 0.001) confirming recent reports that maladaptive coping is more stable than adaptive coping (Aldao, 2013). Maladaptive coping was also more strongly related to well-being (negative relationship) than adaptive coping (positive relationship) both cross-sectionally and longitudinally. Overall, depressive symptoms more highly correlated with dysfunctional coping than paranoia and obsessive-compulsive symptoms. However, results await to be tested in a clinical population. In another study (Dietrichkeit, 2015) the MAX was significantly correlated with the Brief COPE Inventory (Carver, 1997) in depressed patients.

For the present study, we compared a sample of individuals with psychosis to patients with depression and nonclinical controls. Patients with depression can be regarded as an ideal control group for psychosis as – despite high comorbidity (Buckley et al., 2009) – core diagnostic features are different. In contrast, other disorders, especially obsessive-compulsive disorder (OCD), bipolar disorder and anxiety disorders including trauma, show some symptomatic overlap with psychosis as to core symptoms (e.g., a diagnosis of OCD now also includes patients with low insight who sometimes show delusional features; anxiety disorders and psychosis often share great fear; many bipolar patients have delusions). We hypothesized that both individuals with psychosis and individuals with depression would show more dysfunctional and less functional coping than nonclinical controls and that the relationship would be especially pronounced for depressive symptoms (across all populations) against the background that depression has emerged as the strongest correlate of coping in studies (e.g., Moritz et al., 2016). Inconsistent associations across studies for coping behavior and positive symptoms relative to depressive symptoms may, however, be due to the fact that different positive symptoms (e.g., paranoid delusions, grandiosity, hallucinations) are usually pooled to one score (e.g., in the Positive and Negative Syndrome Scale), although there is ample evidence now that the positive syndrome is heterogeneous (Peralta and Cuesta, 1999, Schlier et al., 2015, van der Gaag et al., 2006) and its components may have different etiologies (e.g., Bentall et al., 2014, Zavos et al., 2014). We therefore looked at single positive symptoms more closely by making use of the multidimensional structure identified for the Community Assessment of Psychic Experiences Scale (CAPE; Schlier et al., 2015). No predictions were made whether a distinct profile could be elucidated for patients with psychosis versus those with depression as most studies compared patients with schizophrenia to nonclinical controls only (O'Driscoll et al., 2014).