*3.3. Hierarchical Linear Regression Analyses*

A three-stage hierarchical linear regression analysis was conducted to evaluate the prediction of BAQ scores from the general characteristics of participants (age, gender and education), TAS scores and SOC scores. For the first block analysis, the predictor variables age, gender (coded as 1 = male, 2 = female) and education were analyzed. The results of the first block revealed the model not to be statistically significant (*p* > 0.05). For the second block analysis, the predictor variable TAS scores were added to the analysis and contributed significantly to the regression model, F (4, 92) = 8.501, *p* < 0.001) accounting for 27% of the variation in aggression. Introducing the SOC scores variable at stage three explained an additional 17.8% of variation in aggression and this change in R2 was significant, F (5, 91) = 14.757, *p* < 0.001. Both the TAS scores and SOC scores were significant predictors of aggression. Together, the two independent variables accounted for 44.8% of the variance in aggression. Participants' predicted aggression was equal to 37.84 + 0.108 (TAS) − 0.272 (SOC) (Table 4).


**Table 4.** Summary of hierarchical regression analysis for variables predicting aggression (BAQ scores).

Correlations are statistically significant at the \* *p* < 0.01 level. Beta = standardized regression coefficient.

Two-stage hierarchical linear regression analyses were conducted to evaluate the prediction of BAQ and subscales scores from the scores on the TAS and SOC subscales. For the first block analysis, the predictor variables were the scores on TAS subscales. and for the second block analysis, the scores on the SOC subscales were added. In these analyses, DIF, SOC A, SOC B and SOC C predicted the dependent variables (Table 5).

**Table 5.** Hierarchical regression analyses for variables (TAS and SOC subscales) predicting aggression (BAQ and subscales).


Correlations are statistically significant at the \* *p* < 0.05 or \*\* *p* < 0.01 level. Abbreviations: PA, physical aggression; VA, verbal aggression; H, hostility; A, anger; DIF, difficulty identifying feelings; SOC A; meaningfulness; SOC B, comprehensibility; SOC C, manageability. Notes: 1. Results are given for Step 2, when SOC variables are included as independent variables. 2. Only the variables that significantly predicted the dependent variables are shown.

#### *3.4. Simple Mediation Analyses*

To clarify the nature of the relationship between alexithymia and aggression and answer one of the research questions, we investigated the underlying mechanism by which alexithymia influences aggression through sense of coherence. The objective was to examine the impact of alexithymia on aggression as mediated by sense of coherence. It was hypothesized that being alexithymic will positively predict aggression. Additionally, it was hypothesized that sense of coherence will mediate this relationship. A simple mediation analysis, using the bootstrap method, was conducted to test these hypotheses. Analyzing the indirect effects, the results reveal that sense of coherence significantly mediated the relationship between alexithymia and aggression [(B 0.2093, 95% C.I. (0.1234, 0.2966), *p* < 0.05, Table 6]. Sense of coherence accounted for 65.55% of total effect. These findings provide some evidence that alexithymic patients are less likely to exhibit aggression provided they have high sense of coherence. Nevertheless, alexithymia still contributes to aggression beyond what is accounted for by sense of coherence. Standardized coefficients for the variables are depicted in Figure 1.

**Table 6.** Mediation analysis of sense of coherence (SOC) on Toronto alexithymia scale (TAS) – brief aggression questionnaire (BAQ) relationship.


\* Based on 5000 bootstrap samples.

A bootstrap approach was used to test the significance of the indirect effect of hostility on anger through the mediating role of difficulty identifying feelings. Results of this mediation analysis are displayed in Table 7 and illustrated in the Figure 2.


**Table 7.** Mediation analysis of difficulty identifying feelings (DIF) on hostility (H) – anger (A) relationship.

\* Based on 5000 bootstrap samples. The model explains 60.8% of the variance in the outcome variable.

Results of the mediation analyses to test the significance of the indirect effects of hostility and anger on physical aggression through comprehensibility and manageability are displayed in Tables 8 and 9 and illustrated in Figures 3 and 4.


**Table 8.** Mediation analysis of comprehensibility (SOC B) on hostility (H)–physical aggression (PA) relationship.

\* Based on 5000 bootstrap samples. The model explains 34.34% of the variance in the outcome variable.

**Table 9.** Mediation analysis of manageability (SOC C) on anger (A)–physical aggression (PA) relationship.


\* Based on 5000 bootstrap samples. The model explains 45% of the variance in the outcome variable.

#### *3.5. Path Analysis*

Path analysis was used to determine the pathways by which the alexithymia and sense of coherence dimensions interact to influence modes of aggression. From regression and mediation analysis results, it was expected that comprehensibility and manageability would exert their protective effects, counteracting physical aggression and anger both directly and indirectly through hostility and the difficulty identifying feelings dimension of alexithymia. A structural model with observed variables was tested using a covariance matrix as input and maximum likelihood estimation. This type of analysis provides a comprehensive picture of the nature of the associations between the predictor and dependent variables of interest. Comprehensibility and manageability served as predictor variables and physical aggression and anger were treated as the dependent variables. Results suggested that the measures of fit of the model were satisfactory, indicating adequate fit (*p* = 0.325, CMIN/df = 1.156, comparative fit index CFI = 0.995, normed fit index NFI = 0.968, parsimonious normed fit index PNFI = 0.552, incremental fit index IFI = 0.996, Tucker–Lewis index TLI = 0.990 and the root mean square error of approximation RMSEA = 0.040). No modifications were necessary for the model. The nonexistence of an arrow between two variables means that these two variables are not significantly related. Figure 5 presents the set of hypotheses about the relations between the aforementioned variables.

#### **4. Discussion**

The present study confirmed high levels of aggression and alexithymia and low levels of sense of coherence among schizophrenic outpatients. Participants with alexithymic characteristics reported higher aggressive tendencies and lower sense of coherence capacities. The observed correlations supported all the main assumptions of the relationships between the study variables. Alexithymia and sense of coherence accounted for 44.8% of the variance

in disclosed aggression. The difficulty identifying feelings dimension of alexithymia and the comprehensibility and manageability components of sense of coherence significantly predicted anger, hostility and physical aggression. Sense of coherence mediated the relationship between alexithymia and aggression. The difficulty identifying feelings dimension of alexithymia mediated the relationship between hostility and anger. Comprehensibility mediated the relationship between hostility and physical aggression, and manageability mediated the relationship between anger and physical aggression. From the path analysis, comprehensibility emerged as the key factor counterbalancing alexithymic traits and aggressive behaviors, and manageability effectuated higher anger control. The main hypotheses were supported, indicating that high SOC scores predicted high physical aggression buffering and anger control, both directly and indirectly neutralizing alexithymic traits and hostility.

The presence of hostility in schizophrenic patients is not limited to the acute phase of the disease [109] but persists for a long time after hospitalization [110] and may be a predisposing factor for the emergence of anger and verbal or physical aggression. As opposed to enduring hostility, anger is a temporary but highly intense negative emotional state that usually abates but easily relapses in people who are temperamentally characterized by hostility due to their increased susceptibility to situations of anger [111]. Imaging studies aiming to clarify the neuroanatomical basis of hostility-related dimensions in schizophrenia patients who exhibited high levels of urgency, impulsivity and aggressiveness reported dysfunction of neuronal circuits involving the amygdala, striatum, prefrontal cortex, anterior cingulate cortex, insula and hippocampus [112].

The literature suggests that aggression and alexithymia are related to each other in the mentally ill [21]. Poor emotional awareness [33] and emotion dysregulation [113] are possible implicating mechanisms underlying aggression. By definition, alexithymia is considered to be a disorder of affect regulation [113], and there is evidence that some facets of emotion regulation may be disrupted in schizophrenia [114–116]. A possible etiological sequence is that limited emotional awareness interferes with emotion regulation, effectuating indirect consequences as to emotional response, thereby mediating aggression [117].

Awareness and understanding of emotions are common features of both alexithymia and emotion regulation. Research reports that alexithymics have a variety of emotion regulation difficulties and outline the nature of emotional dysfunction in alexithymia. Nonacceptance of emotional reactions, lack of emotional clarity, difficulties with goaldirected behavior and impulse control, as well as limited access to emotion regulation strategies commonly present in alexithymic individuals [113]. The similarity of content and underlying processes of difficulty in identifying feelings (a dimension of alexithymia) and lack of emotional awareness (a dimension of emotion regulation difficulties) forms a common substrate for both constructs and may be a region of overlap. The observed emotion regulation difficulties in alexithymic individuals appear to be conceptually attributed to a lack of emotional awareness and differentiation [118].

Consistent with results from other studies, it was predominantly the difficulties with identifying feelings (DIF) aspect of alexithymia that was related to aggression [119–121]. Additionally, DIF significantly predicted total BAQ and anger and mediated the relationship between hostility and anger. In other words, the relationship between the cognitive and the affective components of aggression is mediated by the difficulties with identifying feelings dimension of alexithymia. Among those, having difficulty in identifying their feelings of hostile aggression motivated by anger [122] may be considered as a further effort to distract from feelings (or to express feelings in a rather maladaptive way).

Conscious awareness of emotions is acquired when emotional information and experiences are integrated into cognitive processes [123]. Furthermore, it has been argued that many symptoms typical of schizophrenia can be explained by specific cognitive deficits in schizophrenic patients to accurately attribute mental states to themselves or others [124]. Affective theory of mind refers to the ability to comprehend and confront others' affective

mental states and differences in "affective" versus "non-affective" theory of mind' tasks have been reported to relate to certain behaviors in schizophrenia, such as violence [125].

The drastic changes in the subjective experiences of schizophrenic patients give rise to stress and bring about everyday challenges. The ability to manage stressful situations, mental resilience and sense of coherence predict and modify the level of psychological wellbeing among patients suffering from mental disorders [69]. The theory of salutogenesis supports the idea that sense of coherence mitigates stress, and results from early studies evidence the protective effects of a salutogenic approach in individuals with serious mental illnesses [126].

The first step when dealing with a stressful situation is to perceive and be knowledgeably aware of its full extent [53]. An individual with higher comprehensibility is more likely to perceive stimuli from the environment as coherent and understandable [127]. Schizophrenic patients usually require support in reassembling their erratic experiences, reflecting upon them and possibly learning something from them. The severity of symptoms and the chronicity of the psychiatric disorder seem to determine their ability to adapt, as does the use of available resources to manage stressors. First-episode psychotic patients in remission appear to be in a more favorable position compared to chronic patients with persistent positive symptoms or in a deficit state [128,129].

In mental health settings where the therapeutic treatment and long-term follow-up of psychiatric patients is sought, the theory of salutogenesis may find practical application through the modification of therapeutic procedures and environmental factors in order to enhance the three components of salutogenesis [130]. The same approach could even be applied in clinical practice in situations involving conflict prevention through communication, emotional regulation, self-management and other de-escalation techniques [131]. The first goal attempted in de-escalation is reducing the patient's level of arousal to enable discussion by gathering the necessary resources through manageability. The patient should be encouraged to communicate openly with staff about their own emotions and to discuss feelings of anger and frustration in order to enhance comprehensibility. Recent evidence supports the effectiveness of integrating milieu therapy in psychiatric acute wards in reducing conflict behaviors among schizophrenia patients [132].

Preventing or managing stress is of paramount importance according to the diathesis– stress model of schizophrenia. Considering that sense of coherence essentially protects against the pressures generated by stressors, should clearly define the purpose of implemented healthcare interventions, in order to avoid negative health outcomes. Since patients with schizophrenia are so vulnerable to the impacts of stressors, the focus of comprehensive psychoeducation should be on reinforcing coping skills by teaching stress reduction techniques and assertiveness training. Additionally, providing schizophrenic patients with education regarding their illness and improving attention along with other cognitive functions would enhance comprehensibility. When facing challenges, equally important for these patients is acknowledging, approaching, and activating or resorting to other available resources, namely family, social and healthcare support.

In summary, it is reasonable to assume that the inability to recognize and describe emotions may, under stressful circumstances, lead to an increased state of unmanageable arousal. Patients with schizophrenia, particularly if they are characterized by temperamental hostility, feel threatened under stressful conditions and tend to react aggressively. Furthermore, in individuals with reduced abilities to regulate emotions with conscious effort or a strong tendency toward impulsivity, the effect is reinforcing. In these cases, the explanation for the apparent aggressive behavior is provided by the indirect effect of alexithymia, which reduces the cognitive and emotional capacities necessary to moderate distressing feelings and inhibit impulsive actions.

Specific therapeutic interventions to improve the ability to identify subjective feelings that target alexithymia and neurocognitive impairments that may make self-reflection difficult could be implemented. Integrative psychotherapy that targets metacognition could assist patients in developing these capacities [133], and cognitive remediation [134] could provide the prerequisites for metacognitively focused psychotherapy to be successful. Additionally, research argues that alexithymic individuals would benefit more from group-based psychological therapy and supportive and educational approaches as opposed to interpretive ones [25]. A recent open-label randomized controlled trial applied an integrated cognitive-based violence intervention program on management of repetitive violence in patients with schizophrenia and evidenced significant improvement of cognitive failure, management of alexithymic features and attribution styles and errors and fostered adequate decision-making styles and emotion regulation capacity. This intervention provided patients a more active role to manage their violent behavior with the involvement of alexithymia [75].

In the last two decades, there has been increasing research interest in the ability of tailored interventions to modify and strengthen the sense of coherence of various target groups [97]. Enhancing comprehensibility for individuals with schizophrenia should be the focus of comprehensive psychoeducation and cognitive remediation. Moreover, aiming to improve prognosis through the implementation of psychosocial rehabilitation interventions requires the identification and reduction of aggravating factors, such as ineffective stress management, as well as the enhancement of protective factors, especially manageability.

Aggressive behavior is a leading public health problem incurring a massive cost on society and, according to epidemiological studies, individuals diagnosed with major mental disorders such as schizophrenia are more likely to be engaged in such behaviors than the general population, with obvious relevance for health care systems. The etiological heterogeneity of aggression and the possible multifactorial contributors to aggressive behaviors, along with the fact that the current treatment approaches and outcomes are so far inferior, justifies the search for new targets for the treatment of aggression in people with schizophrenia [135]. Addressing factors that limit the effectiveness of treatments might decrease the burden of these severe chronic disorders. Given the high observed prevalence of alexithymia and associated negative outcomes, researchers and clinicians examined the feasibility of treating alexithymia [136]. Interventions aiming to increase patients' emotional awareness and their ability to label emotions may promote their successful engagement in cognitive behavioral psychotherapies to regulate unpleasant, angry emotions before they escalate and drive their behavior [137]. There is promising evidence that alexithymia can improve with treatment even after neurological damage [32].

Identified targets for treating aggression in schizophrenia in our study, i.e., alexithymia and sense of coherence, are mostly amenable to psychotherapeutic and psychosocial interventions. Several different psychotherapeutic approaches for schizophrenia have been developed and studied, with cognitive behavior therapy having the strongest evidence base [138]. Providing comprehensive psychological interventions in this clinical population will likely require drawing upon knowledge from several areas of current research and incorporating elements of various psychosocial interventions, such as cognitive remediation, social skills training and psychoeducation [139]. A recent systematic review evidenced the effectiveness of cognitive remediation and social cognitive training in the management of violent and aggressive behaviors in schizophrenia [140]. According to another review, once patients' positive symptoms have stabilized, cognitive behavioral therapy and cognitive remediation are the two psychosocial interventions that have demonstrated positive outcomes for violence in patients with schizophrenia [141].

The presence of contributing factors to aggression, either adverse or protective, namely alexithymia and sense of coherence, acquires importance in light of the possibility of handling their impact on the effectiveness of antiaggressive treatments, particularly in the domain of psychosocial interventions. Testing the efficacy of psychotherapeutic interventions for people with schizophrenia may benefit from an inclusion of aggression as a treatment outcome in clinical trials. Our study proposes alexithymia and sense of coherence as putative targets for addressing aggression in schizophrenia and extends ideas for treatment and research. Future studies should be carried out, especially controlled and follow-up studies, comparing different forms of treatment on more extensive patient

populations while considering potential confounding factors and performing objective assessment of aggression, alexithymia and related constructs.

The present study suffers from various limitations. Participants were enrolled from the outpatients department. Subsequently, research findings may not be applied beyond this study population. Additionally, since the data were collected using self-report tools, self-serving bias may be an issue. Specifically, the TAS-20 scale as a measure of alexithymia consists of self-rated agreement statements and presupposes awareness of the deficit to be reported, which raises the concern that emotion recognition deficits that patients do not detect would not be captured by this scale. Moreover, due to social desirability bias, participants might have stated their self-reported levels of aggression in a socially acceptable manner instead of providing answers that are reflective of their genuine aggression level. Finally, the cross-sectional design of the study precluded us from making secure inferences about direction of causality.

#### **5. Conclusions**

High rates of aggression and alexithymia, along with low sense of coherence, were observed among schizophrenic outpatients. Alexithymia fueled aggression, and sense of coherence counteracted aggressive tendencies. The difficulty identifying feelings dimension of alexithymia emerged as a liability, and the comprehensibility component of sense of coherence as a protective factor buffering the deleterious consequences towards physical aggression. These results hold practical implications for the treatment and rehabilitation of schizophrenic patients.

**Author Contributions:** Conceptualization, A.P. and A.T.; methodology, A.T., I.I. and A.P.; software, A.T., I.I. and A.P.; validation, A.P., A.T. and S.B.; formal analysis, A.P., A.T. and E.T.; investigation, A.P., S.M.P. and K.G.; resources, A.P.; data curation, S.M.P., E.T. and A.P.; writing—original draft preparation, A.T., A.P. and E.K.; writing—review and editing, A.T. and A.P.; supervision, A.T. and A.P.; project administration, A.T. and A.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Institutional Review Board Statement:** The study was conducted in accordance with the Declaration of Helsinki and approved by the Clinical Research Ethics Committee of Sotiria General Hospital (Number 24252/27-9-21).

**Informed Consent Statement:** Informed consent was obtained from all subjects involved in the study. Participation in the research was voluntary.

**Data Availability Statement:** The data and the questionnaires of the study are available upon request from the corresponding author.

**Acknowledgments:** We would like to thank all participants in our study.

**Conflicts of Interest:** The authors declare no conflict of interest.

### **References**

