Mentalization in young patients undergoing opioid agonist treatment: Implications for clinical management

Highlights • Mentalization among opioid addiction (OA) patients differs from normative data.• Younger patients and with the most severe disorder have higher uncertainty score.• Patients with a more severe abuse have lower certainty score.• Patients with previous pediatric mental-health contacts have lower certainty score.• Patients receiving a therapeutic community support have lower certainty score.


Introduction
Substance Use Disorders (SUDs) have their onset before age 25 in 49 % of patients, and before age 18 in 15 % of them (Solmi et al., 2022), presenting with a high long-term clinical impact (Degenhardt and Hall, 2012). Among them, opioid addiction (OA) results in high rates of disability (Lusk and Stipp, 2018) and mortality (Hser et al., 2015). While Opioid Agonist Treatment (OAT; i.e., the controlled administration of opioid agonist drugs, with the main aim of reducing craving for the substance of abuse) has proven to be an effective treatment in controlling the disorder and reducing mortality risk (Hser et al., 2015;Santo et al., 2021), poor treatment adherence, post-treatment relapses, and risk behaviors remain unmet clinical needs (Krawczyk et al., 2020;Pearce et al., 2020), urging a better understanding of OA determinants to improve outcome.
Independently of biological susceptibilities (Deak and Johnson, 2021;Patriquin et al., 2015;Volkow et al., 2016), psychosocial risk factors have been suggested to contribute to the development of a SUD. A number of psychosocial factors have been explored, including, but not limited to, poor self-regulation, poor decision-making skills, and insecure attachment style (Gerra et al., 2021;Mishra et al., 2022;van Draanen and Aneshensel, 2022;Vungkhanching et al., 2004). Such evidence emphasizes the important role that interpersonal relationship and, more in general, social difficulties may play in sustaining SUD and OA.
Known for more than 30 years, the term mentalization refers to the ability to use internal mental states to manage and understand one's own and others' behavior (Fonagy and Luyten, 2009) at both the emotional and cognitive levels (Luyten et al., 2020). Generally, both excessively low use of mentalization ('hypo-mentalization', 'uncertainty' about mental states) and excessively high use ('hyper-mentalization', 'certainty' about mental states) are considered possibly dysfunctional. Mentalization is currently receiving much attention among studies because of the role that psychosocial determinants, such as interpersonal relationships, may have in altering its acquisition (Luyten et al., 2020), with implications for the development of mental disorders (Luyten et al., 2020). Indeed, among individuals exposed to dysfunctional social environments or traumatic events, especially in early childhood, the development of mentalization may be inhibited (Borelli et al., 2019). As a consequence, such individuals would present with a socially acquired (Sng et al., 2018) non-mentalistic interpretation of stimuli, based on representation of others' behavior in terms of observable contents of experience (Gergely, 2003). The inability to process the cognitive and emotional components of social experience would lead to externalizing symptoms, such as impulsivity, inattention, interpersonal problems, and poor academic maturation Luyten and Fonagy, 2018;Perroud et al., 2017) as well as substance abuse  and other addiction symptoms such as gambling (Cosenza et al., 2019), in a dysfunctional attempt to cope with stress and reduce arousal (Luyten et al., 2020).
To the best of our knowledge, studies investigating mentalization in SUDs, especially in patients with OA undergoing OAT, are lacking. The aim of this study was twofold: (i) to describe mentalizing abilities among OA patients undergoing OAT; (ii) to investigate the association between mentalization characteristics on one hand, and sociodemographic and clinical characteristics on the other, in this specific patient population. Mentalization impairments in patients receiving OAT were hypothesized, along with differential sociodemographic and clinical profiles as a function of mentalization abilities.

Study design and sample
This study was conducted at the Drug Addiction Service of Friuli Centrale Health University Authority, Udine, Italy, an outpatient facility specifically devoted to the care of people experiencing SUDs. All young patients aged 18-30 years old, consecutively assessed for OA and treated with OAT over the period July to November 2021, were included into the study. Patients who were deemed clinically unstable (e.g., treatment stabilization not reached yet) as well as with a known or suspected acute substance intoxication were excluded.
The authors assert that the work described here has been carried out in accordance with The Code of Ethics of the World Medical Association (Declaration of Helsinki) for experiments involving humans as well as the Uniform Requirements for manuscripts submitted to biomedical journals.

Assessment
The assessment took place during the scheduled appointments. It included: (i) a general data collection form, with additional relevant information about the study participants being extracted from the clinical records; (ii) structured and semi-structured clinical interviews concerning the SUD-related symptoms and any other psychiatric symptoms/disorders; iii) a self-administered evaluation of mentalizing abilities. A brief description of adopted instruments is provided below.
MINI-7. The official Italian version of the Mini International Neuropsychiatric Interview, 7th edition (MINI-7), was used (Sheehan et al., 1998;Sheehan, 2016). It is a semi-structured interview assessing 17 psychiatric disorders according to DSM-5 diagnostic criteria.
RFQ. The Reflective Functioning Questionnaire (RFQ) (Fonagy et al., 2016), in its official Italian version, was used. RFQ studies conducted among Italian adults support its structure validity (Morandotti et al., 2018) and psychometric quality (Velotti et al., 2021). It consists of eight items scored on a 7-point Likert's scale (from 'Completely disagree' to 'Completely agree'), measuring two mentalizing dimensions: (i) 'Uncertainty' about mental states (RFQ-U), with high scores indicating hypo-mentalization (i.e., concrete thinking, with difficulties in understanding one's own and others' mental states; ii) 'Certainty' about mental states (RFQ-C), with high scores indicating hyper-mentalization (i.e., a rigid and biased attribution of mental states, that go far beyond available evidence). As an example, a person shows more uncertainty about others' mental states if they agree with the statement: "People's thoughts are a mystery to me". On the other hand, a person stating "I always know what I feel" shows certainty with respect to one's own mental states. The two scales are measured by assigning different weights to partially overlapping items, so that both scales are composed of six items and give a score ranging from 0 to 18.

Statistical analysis
Welch's corrected t-tests were used for group comparisons with homoscedastic measures, and Mann-Whitney's test in case of heteroscedasticity. Associations between continuous measures were analyzed with Pearson's correlations (r), presented with their 95 % confidence intervals (CIs). All bivariate correlations were calculated on 37 observations. Fisher's z tests were used to compare correlation coefficients. In correlation analysis, to control for multiple hypothesis testing, Benjamini-Hochberg's procedure was adopted. Cronbach's αs for the RFQ scales were reported, with their 95 % CIs (calculated by Feldt's method).
In the main analysis, multiple linear regression models were then used on standardized measures to test for any association between variables, by correcting for the effects of potential confounders. Given the small sample size and the preliminary nature of this study, forward selection of predictors informed by Akaike's Information Criterion (AIC) was used. The initial pool of potential measures was chosen based on the statistically significant results of the preliminary analyses. For multiple linear regressions, a post-hoc power analysis was conducted, reporting in main-text only sub-optimal results (i.e., β > 0.20) in terms of achieved power for statistically significant results.

Recruitment process
Fifty-eight patients on OAT aged 18-30 years were considered for inclusion in the study. Of them, two were found not to be eligible, based on their medical records (i.e., one patient with cognitive disability, another patient with severe psychotic disorder). Further, thirteen patients were excluded because they were deemed clinically unstable by the referring clinicians. Finally, during the recruitment process, three patients moved to another location, one patient died, and another discontinued OAT. Informed consent was then offered to 38 patients. During the assessment period, a patient dropped out of the study before concluding all the evaluations, leaving a final sample size consisting of 37 participants.

Mentalizing characteristics
Significant differences were found when comparing the mentalization skills measured in our sample (M ± SD standardized in z-score on normative scores: RFQ-U, +0.70 ± 1.193, t 49.9 = − 3.26, p = 0.002; RFQ-C, − 0.43 ± 0.644, U = 1000.0, p = 0.033) with those referring to available normative data in non-clinical young adults (18-30 years; M ± SD: RFQ-U, 3.57 ± 3.335; RFQ-C, 6.50 ± 4.488) reported by (Morandotti et al., 2018). A graphical representation of the comparison between our sample and the normative sample is provided in Fig. 1 Univariate analyses, indicating associations between sociodemographic and clinical characteristics on one hand, and mentalizing characteristics on the other, are provided in the Supplementary results.

Association between SUD and mentalizing characteristics
Univariate analyses, indicating associations between SUD and mentalizing characteristics, are provided in the Supplementary results and Supplementary Table 3).
The overall final model for the RFQ-C scale was also statistically significant (F 7,30 (Table 3).

Discussion
Findings from this study indicate difficulties in mentalization, in terms of higher uncertainty and lower certainty, among patients receiving OAT for their opioid addiction, compared to the profile of normative data from an age-matched non-clinical sample. Also, higher uncertainty was found among younger patients and in those with the most severe substance use disorder (SUD) in terms of craving and need for care. Finally, lower certainty was found in those with a more severe substance abuse, previous contacts with pediatric mental-health services, and receiving a therapeutic community support.
When exploring characteristics associated with altered mentalization, both direct (e.g., opioid abuse, need for care) and indirect (e.g., previous pediatric mental-health contacts, school failures) psychosocial features among SUD patients were found accompanying altered mentalizing patterns, in line with theoretical models of aberrant mentalization development (Fonagy and Luyten, 2009;Luyten et al., 2020). Specifically, in the context of such psychosocial difficulties, OA patients presented higher scores on the uncertainty scale, that reflects hypomentalization, an interpretation of social experiences mainly based on a non-mentalistic observation of others' behaviors (Gergely, 2003). Further, results from this study confirmed and extended previous evidence of an association between mentalizing difficulties and mental health symptoms more in general, believed to result from an inefficient social processing of cognitive and emotional stimuli Luyten and Fonagy, 2018). In fact, greater hypomentalization was found among SUD patients also suffering of GAD, possibly as a dysfunctional response to stress and hyperarousal states (Borelli et al., 2019;Luyten et al., 2020).
Certainty, whose extreme scores would indicate a dysfunctional hyper-mentalization, i.e., an interpretation of others' behaviors without evidence supporting that presumption, was lower in patients with more severe SUD and related psychosocial disabilities. When assessing the multifaceted components of mentalization, research evidence indicates that uncertainty scores show a positive correlation with alexithymia (i. e., difficulties in identifying, understanding, and describing emotions), and a negative correlation with empathy (i.e., understanding the mental states of others while resonating with them) and mindfulness (i.e., awareness of what one experiences). Instead, certainty scores would show the opposite pattern, correlating negatively with alexithymia and positively with empathy and mindfulness (Badoud et al., 2015). In line with these findings, our results would suggest poor mentalizing abilities, in terms of reduced empathy and mindfulness as well as increased alexithymia, among OA patients presenting with greater abuse, service use, and psychosocial difficulties.
Results from this study must be seen considering its strengths and limitations. To the best of our knowledge, empirical evidence about mentalizing abilities in SUD is lacking, making the present findings valuable for clinicians and researchers working in the field. Also, a clinically well-defined sample was obtained, to limit the implications of "spurious comorbidity", that is the higher co-occurrence of disorders in clinically ascertained samples than in population-based samples, possibly due to patients presenting with multiple conditions being more likely to seek medical care and receive a diagnostic evaluation. However, the limited sample size, obtained in a single recruitment hub, suggests caution in drawing conclusions about mentalizing abilities in SUD, requiring replication of the findings in larger multisite studies. Still referring to the small sample-size, we emphasized that the power calculated a posteriori for some results was sub-optimal (i.e., for some coefficients of the multiple regression models). Also, participants included in the study were patients in frequent contact with clinicians for their therapeutic needs, with regular medical check-ups, and adherence to health-care pathways. While offering advantages in terms of recruitment feasibility and sample homogeneity, this may have similarly affected the generalizability of the results to the wider population of individuals with OA. Also, data collection was carried out during the pandemic restrictions for COVID-19. Although restrictions were reduced during the assessment period, we have no way of assessing whether the reported data may have been influenced by the atypical contextual situation. Finally, the RFQ that was used to assess mentalization characteristics in our sample, has been criticized. Also, in our observation the scales of the RFQ showed poor internal consistency. It has been suggested that, despite reducing administration time, administrator training, and burden on participants (Fonagy et al., 2016), its psychometric reliability may be lower when compared to the Reflective Functioning Interview, the gold standard measure for the assessment of mentalization (Anis et al., 2020;Morandotti et al., 2018). Arguably, at the psychometric level, the RFQ seems to be better suited at capturing a single dimension of dysfunctional mentalization style (Müller et al., 2022). Nevertheless, in our study we preferred to refer to the available Italian norms (i.e., organized in the two scales originally proposed), to allow a preliminary assessment of this poorly investigated area and a comparison with other observations available in the literature. However, these observations need further and deeper investigation, using more reliable assessment instruments, probably based on clinical interviews. In addition to a more reliable assessment of the characteristics of mentalization in patients with OA, future investigations may also clarify any associations with other psychiatric problems in comorbidity. For this, larger sample sizes will be needed to assess interactive effects between different disorders.
In conclusion, this study provides preliminary observations in an under-explored field, that is the evaluation of mentalizing abilities in SUD and its clinical course. Mentalization patterns among OA patients undergoing OAT were found to differ from what expected based on normative data. Also, patients with greater mentalizing dysfunction present with a higher burden in terms of SUD severity, comorbidities, psychosocial disabilities, and service use. Thus, findings may have important public health implications, as they suggest that interventions targeting mentalization may have positive repercussions in preventing SUD, mitigating its severity, and containing its healthcare and social costs.

Contributors
Substantial contributions to the conception or design of the work and/or the acquisition, analysis, or interpretation of data for the work: all authors. Drafting of the manuscript and/or revising it critically for important intellectual content: all authors. Final approval of the version to be published: all authors. Agreement to be accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved: all authors.

Declaration of Competing Interest
The authors declare the following financial interests/personal relationships which may be considered as potential competing interests: M.C. has been a consultant/advisor to GW Pharma Limited, GW Pharma Italy SRL, and F. Hoffmann-La Roche Limited, outside of this work. All the other authors declare no conflict of interest.

Data availability
Data will be made available on request.