Investigating defensive functioning and alexithymia in substance use disorder patients

According to the Diagnostic and Statistical Manual of Mental Disorders, fifth edition (DSM-5) [1], Substance Use Disorder (SUD) “describes a problematic pattern of using alcohol or another substance that results in impairment in daily life or noticeable distress”. Diagnostic criteria are uniform across different drugs of abuse and include common-sense symptoms such as craving, tolerance, withdrawal, failure to fulfill social obligations, and use despite harms.

SUDs cause a great deal of personal suffering for patients and families, and have devastating psychological, medical, and social effects [2]. Despite the prevalence and the numerous disabling outcomes associated with SUD, its risk and protective factors from both the individual and social perspectives are still unclear. Research has mostly investigated personality traits [3], psychopathology [4], and coping style [5], but recent evidence highlights how defense mechanisms and emotion regulation may play a crucial part in the onset and development of this disorder [6, 7].

First described by Sigmund Freud (1894) and later systematized by Anna Freud (1936), defenses represent involuntary cognitive operations that occur outside of awareness to minimize sudden changes in internal and external environments by modifying the conscious experience of thought, feeling, and emotion [8]. The DSM-5 defines defenses as “mechanisms that mediate the individual’s reaction to emotional conflicts and to external stressor. Some defense mechanisms are invariably maladaptive. Others may be either maladaptive or adaptive, depending on their severity, their inflexibility, and the context in which they occur” [1]. Defenses could be arranged according to the Vaillant continuum ranging from being maladaptive to being adaptive [8].

Findings from several studies suggest that defenses are related to psychological health and well-being, psychiatric symptoms, and some evidence suggests that defenses may antedate symptoms and predispose to psychopathology [9]. Defensive functioning is found to be associated with personality disorders [10], and several emotional problems [11]. Maladaptive defenses are strongly associated with several other indices of poorer mental health and relatedness, encompassing alexithymia [12], insecure attachment [13, 14], distancing/avoidance coping strategies [15], deliberate self-harm [16], and misperception of mortal risks [17]. Adaptive defenses tend to be associated with better health conditions [18].

Among the few studies available on defense functioning in SUDs, substance-dependent individuals use rationalization, projection, denial, and suppression defenses more than healthy individuals [19]. Studies have shown that SUD is associated with more frequent use of immature defense mechanisms such as autistic fantasy, acting out and isolation [5]. Consistently, in young alcohol dependent individuals, acting out was associated with a higher risk of self-injury, while less frequent use of mature defense mechanism seems to be a predictor of suicidal behavior in the same population [20]. Acting out defense has also been previously associated with relapse to alcohol use 12 months after inpatient treatment [21]. Notably, previous studies also highlighted the association between maladaptive defense mechanisms and antisocial behavior in alcohol dependent patients [22]. This is consistent with other studies in which substance dependent patients exhibited a significant association between the use of immature defenses and severity of dependency, dissociative experiences, and alexithymia [23].

Indeed, there is a high prevalence of alexithymia in patients with SUDs [24].

Used for the first time by Sifneos [25] to describe certain clinical characteristics observed among psychosomatic patients, the term alexithymia refers to a multidimensional personality construct, defined by a set of four characteristics: 1) difficulty in identifying feelings and in distinguishing feelings from bodily sensations of emotional arousal, 2) difficulty in describing and in communicating feelings to others, 3) lack of fantasy and imagination, and 4) an externally oriented style of thinking [25]. The issue about the nature of the construct, i.e., whether it is more a deficit in emotional processing, or a defensive process, has triggered a vivid debate among psychoanalysts [26]. Even if the deficit view has gained support from experimental research [27], other views have suggested moving beyond the debate by giving a teleological sense to the term “defense”. Indeed, according to McDougall [28], these competing theories about the construct are not mutually exclusive: alexithymia is indeed related to deficits in the cognitive representation of emotions but could also be conceptualized as a massive defense against primitive terrors, rather than neurotic conflicts Several studies have reported high rates of alexithymia in drug-dependent individuals [29] and a significant positive association was found between alexithymic traits and craving, the severity of the disorders, and related difficulties [30].

Alexithymia has a complex relationship with various risk factors for the development of SUD related problems (i.e., drug expectations, negative affectivity, insecure attachment, and personality disorders) [29]. Furthermore, the association between immature defensive patterns and alexithymia is well known [12]. In previous studies considering the relationship between alexithymia and defensive array in alcohol dependent patients [23], neurotic and immature defense styles (assessed with DSQ) were higher in the alexithymic group than the non-alexithymic group, and alexithymia positively correlated with the use of passive aggression, acting out, isolation, fantasy, denial, and total immature defense score.

However, it should be noted that the great majority of studies investigating the association between defensive functioning and alexithymia in SUD has relied upon the self-report Defense Style Questionnaire (DSQ) [31]. DSQ, despite methodological challenges in studying unconscious processes by self-report, is the most widely used instrument for defense measurement with validated versions in numerous languages and it showed good reliability and validity [32]. DSQ contains some psychometric shortcomings regarding the overtly pathological wording of the self-report and the unstable factor structure that partially differs from the defensive array described by Bond et al. [31]. To overcome some of the psychometric limitations of the DSQ, the Response Evaluation Measure (REM-71) has been developed [33]. The REM-71 has some advantages when compared with the DSQ. REM-71 items are simpler than DSQ items; they avoid the overtly pathological wording of some DSQ items; phrases involving outcomes (I do X to achive Y) are not used to avoid the confound of dependent and independent variables; REM-71 has a coherent two-factor structure; factor scales are reliable [33]. This questionnaire defines defenses as normative and ubiquitous self-regulatory processes, in line with Vaillant’s model [8].

To the best of our knowledge, no study has been so far conducted using the REM-71 with the specific aim of investigating the relationship between defensive functioning and alexithymia. Therefore, the goal of this study was, first, to investigate potential differences in terms of defensive functioning between SUD patients and non-clinical controls. Secondly, we aimed at investigating the relationships between alexithymia and maladaptive/assimilation defenses (REM-71 Factor 1) in the SUD group. As a last step, we aimed at characterizing in the sample of SUD patients the potential association between defense type (i.e., REM-71 Factor 1: maladaptive/assimilation defenses; Factor 2: adaptive/accommodation defenses) and personality and psychopathological characteristics. We expected that SUD patients would show a more maladaptive defensive functioning compared to that of controls.

Furthermore, we expected that a pattern of maladaptive defenses in the clinical sample would show strong correlations with alexithymia scores, and that an above-cutoff score in REM-71 Factor 1 (4.40) [9] would be associated with impaired psychological well-being, and with difficulties in identifying and describing feelings.

One hundred and seventy-one first-time admitted patients were recruited in nine therapeutic communities in southern Italy, Department of Addiction Services of Bari, National Health Service (ASL BA).

The criteria for inclusion in the study were the presence of a formal diagnosis of SUD (according to the DSM-5), and age above 18. Patients with independent or substance-related current major psychiatric disorders were excluded after a comprehensive psychiatric assessment. Clinical participants had been undergoing their clinical treatment for not less than 1 month and not more than 3 months. One hundred and forty-two (83%) were males and 29 (17%) were females. Their mean age was 36.5 years (SD = ± 8.41).

Among SUD patients, 52 (30.4%) had a diagnosis of Alcohol Use Disorder from moderate to severe, 35 (20.5%) had a diagnosis of Stimulant Use Disorder (cocaine) from moderate to severe, 84 (49.1%) had a diagnosis of Opioid Use Disorder (heroin) from moderate to severe. Ten (5.8%) had a primary school diploma, 99 (57.9%) had a first-level secondary school diploma, 54 (31.6%) had a second-level secondary school diploma, eight (4.7%) had a bachelor degree. Participants were monosubstance-dependent users (with the exception of nicotine, caffeine, and/or past cannabis dependence). They were recruited after the toxicology screening procedure that was independently performed by the therapeutic communities.

One hundred and fifty-five control participants with no history of SUDs were recruited randomly in the population through advertisement. Controls were screened with the Alcohol Use Disorders Identification Test (AUDIT) [34], the Drug Use Disorders Identification Test (DUDIT) [35], and the Italian version of the Structured Clinical Interview for DSM-5 (SCID-5.CV) [36]. Controls with psychiatric disorders were excluded. One hundred and thirty (83.9%) were males and 25 (16.1%) females. Their mean age was 37.6 years (SD = ± 9.94). 6 (3.9%) had a primary school diploma, 74 (47.7%) had a first-level secondary school diploma, 64 (41.3%) had a second-level secondary school diploma, 11 (7.1%) had a bachelor degree. The study was approved by the Institutional Review Board of the Department of Addiction Services, National Health Service (ASL Bari), and it was conducted in accordance with the Declaration of Helsinki.

All subjects provided their written informed consent.

In this study we assessed defensive functioning in SUD patients using the REM-71. The first aim of the study was to investigate the defensive functioning in SUD patients compared to the defensive functioning of non-clinical controls.

Our results support the hypothesis that SUD patients show a more maladaptive defensive pattern than that of non-clinical subjects. When considering the SUD group, we found a significantly higher scores for 5 out of 14 Factor 1 defenses than that of the control group (Acting out, Fantasy, Omnipotence, Projection, Undoing). Among Factor 2 defenses, no differences were found except for the higher score on Intellectualization in the control group and higher Sublimation scores in the clinical sample. SUD patients with a F1 score above the clinical threshold showed a worse psychological functioning in all the domains explored in the study, compared to SUD patients with a F1 score under the cut-off.

When compared to controls, Stimulants and Opioids Use Disorder patients shared the same defense pattern characterized by the use of Acting out, Fantasy, and Sublimation. No differences in the MMPI (with the exception of TPA higher scores in Stimulants Use Disorder patients), and alexithymia scores where found between Stimulants and Opioids Use Disorder patients. In the only previous study using the DSQ with opioids dependents [40], heroin dependent inpatients were using immature defense mechanisms (particularly acting out and splitting) than the controls.

In our study, Alcohol Use Disorder (AUD) patients showed a more maladaptive array of defense, including Acting out, Fantasy, Omnipotence, Projection, Undoing. These results are in line with previous studies [5, 23]. Because immature defenses generally block conscious awareness of distressing material, operating a rigid and excessive distortion or reality rather than allowing the individual to consciously acknowledge it [41], using substances may be a chemical way of coping with problems. AUD and SUDs patients showed higher levels in Sublimation. Notably, the use of sublimation among addicted patients is not so rare. In a study conducted in Turkey, sublimation, pseudoaltruism, acting-out, isolation and autistic fantasy discriminated substance dependents from healthy controls [22]. Grebot and Dardard [42] also found a significant association between the intensity of cannabis addiction and the use of sublimation. In a study by Abd Halim and Farhana Sabri [43], sublimation was a prominent defense in relapsing addicts. Sublimation is a constructive defense employed to deal with unacceptable thoughts, emotions, and impulses, by channeling them into positive and socially acceptable behaviors. The urge to use potentially lethal drugs is an expression of hostility towards those around the addicted person. Family members cry and worry that they will find their loved one dead. There is something wrong with our finding that drug use represents a “sublimation”. Drug use is not a “positive and socially acceptable behavior”. Therefore, we must call our finding of more “sublimation” in drug users over controls a construct error in the research design. It is probably impossible to use drugs addictively while employing a mature defense such as sublimation that makes loving those around one at least partially conscious. Therefore, our finding could be a meaningless artefact probably due to the low α value of the REM-71 Sublimation domain (α = .22).

In our study, with regard to personality and psychopatological differences among SUD sub-groups, AUD patients showed a pattern of depressive symptoms when compared to Stimulant Use Disorder patients. Epidemiological data suggest that the linkages between the AUD and major depression (MD) cannot be accounted for fully by common factors that influence both disorders, and that the disorders appear to be linked in a causal manner. According to metanalytic data [44], the most plausible causal association between AUD and MD is one in which AUD increases the risk of MD. Potential mechanisms underlying these causal linkages include neurophysiological and metabolic changes resulting from exposure to alcohol. Furthermore, the use of maladaptive defenses among dependent patients might be the consequence of alcohol dependency during the active phase of disorder. The capacity to use adaptive defenses may diminish in the acute phase of the disorder and AUD patients may use more maladaptive defenses (acting out, projection, dissociation, splitting). During sobriety period, as the depressive and anxious symptoms remitted, their defensive functioning may return to a higher level of maturity [45].

Future longitudinal studies should investigate the changes in defense functioning during detoxification.

No significant gender differences were found in the use of defenses among control participants, with the exception of higher scores on REM-71 Somatization. Female participants tend to deal with emotional conflicts or internal/external stressors by the expression of psychological conflict via bodily symptoms without symbolic content. This finding is in line with the previous research that observed somatization as far more common in women [46‐48]. Research findings showed that women tend to use more internalizing defenses (i.e. somatization), while men tend to use more externalizing defenses (i.e. acting out), in line with theoretical and clinical observations [49]. However, some studies on gender difference in defense style have provided conflicting findings [32], pointing out the role of cross-cultural factors reinforcing or prohibiting the use of certain defenses via socialization patterns [50]. In the SUD group, there were important gender differences in the frequency of use of Factor 1 defenses and altruism. Watson and Sinha [50] found that female subjects are more likely to use altruism, consistent with the results of our study.

The defenses endorsed more strongly by female participants were Acting out, Splitting, Displacement, Dissociation, Fantasy, Projection, Repression, and Somatization.

In our study, we found strong correlations between TAS-20 scores and the use of maladaptive/assimilation defenses. Additional to the previous studies with the DSQ we found the TAS-20-DIF to be the main alexithymia factor related to Factor 1 defenses. Several studies suggest that individuals with higher DIF scores, conversely with respect to EOT, showed higher psychiatric symptoms and distress [51]. Consistently, Ueno et al. [52] conducted a cluster analysis on healthy individuals and found two distinct alexithymia subtypes characterized, respectively, by higher scores in DIF and neuroticism, and by higher scores in EOT and lower openness to experience. Moreover, Chen and colleagues [53] found that individuals with introversive-high-alexithymia (characterized by higher scores in DIF and DDF, and lower scores in EOT) tended to show fewer effective strategies to regulate their emotions.

A regression analysis showed that Projection, a psychological defense mechanism proposed by Anna Freud (1937) in which an individual attributes unwanted thoughts, feelings and motives onto another person, explained the 38% of the variance of DIF, after controlling for gender, age, and years of education. In our study, DDF was mainly predicted by Repression at 28%, whereas EOT was independently predicted by Displacement (9%), and it was weakly associated to maladaptive/assimilation defenses. Focusing on the association between Factor 1 defenses and alexithymia, it is noteworthy that DDF and EOT mean scores were explained by more hysterical/neurotic (Repression and Displacement) defenses, according to the Vaillant hierarchical classification [8]. On the other hand, DIF was mainly associated with Projection, a defense characterized by disavowal of reality. Therefore, individuals with higher impairments in the ability of identifying feelings, and in the modulation of affective processes via cognitive strategies, tend to apply maladaptive strategies (including more maladaptive defenses) for the regulation of affect in stressful situations [54].

In SUD group, Factor 1 and Factor 2 scores shared positive correlations with 10 out of 29 clinical variables. All the statistically significant Factor 2 correlations had values going from small to moderate effect size. Factor 2 scores did not correlate (or showed small negative correlations) with most of the clinical scales of MMPI-II (with the exception of Schizophrenia and Hypomania), and with the content scales designed to assess anxiety and depression. Factor 1 scores showed positive correlations with 27 out of 29 clinical variables with moderate to strong effect size. These results were generally supportive of the expected relationships between maladaptive/assimilation defenses and psychopathological distress across several domain. On the other hand, Factor 2 in our clinical sample showed less informative results and did not clearly correlate with a better functioning. These findings need further exploration and research to shed light on the role of Factor 2 defenses in addiction.

There are important limitations to this study. First, defense style was assessed through self-report instruments. Criticism has been raised against the use of self-report for the assessment of defensive functioning [55]. Future studies might address this issue by comparing the REM-71 scores with clinical ratings, with the Thematic Apperception Test (TAT) [56] according to Cramer’s instructions [57], and with performance-based tests (i.e., Rorschach R-PAS [58]). Furthermore, two defenses, Sublimation (α = .22) and Denial (α = .25), had α values less than .40. These findings suggest a need for a reconsideration of this two REM-71 subscales in the future.

The cross-sectional design of the study does not allow any inference on the causal relationship between defenses and the other psychological variables assessed in this study. Prospective and longitudinal studies, according to Bond suggestion [59], would be the ideal study to find correlations between preexisting defenses and specific illness.

Furthermore, these findings provide support of the adoption of the REM-71 as a useful screening instrument among addicted patients.