Personality in Alcohol Use Disorder: A Study Focused on Sex Differences

1. Introduction

Alcohol use disorder (AUD) is characterized by a maladaptive pattern of alcohol consumption resulting in clinically significant impairment or distress [1]. AUD and mental disorders, including Personality Disorders (PDs), are commonly associated [2]. Lifetime prevalence of AUD in major mental disorders is 22.3% while in subjects with no mental disorder are 11.0% [3]. Epidemiological data indicate that approximately 28.6% of individuals with AUD meet criteria for at least one PD [4]. Estimates of PD prevalence among AUD populations varies widely from 13.9% [5] to over 48% in some clinical samples [6].

The study of prevalence of PDs in AUD subjects have been carried out using different instrument to evaluate personality such as the Multiphasic Minnesota Personality Inventory (MMPI) [7], the NEO personality inventory [8], the Eysenck Personality Questionnaire (EPQ) [9], the Structured Clinical Interview for DSM Personality Disorders (SCID) (Balachandran et al., 2023) and the Cloninger Temperament and Character Inventory [11] just to name some of them.

The Millon Clinical Multiaxial Inventory (MCMI) distinguishes personality pathology from clinical symptoms and aligns its scales with the DSM-5 PDs classification, grouping personality types into three clusters: A (schizoid, schizotypal, paranoid), B (borderline, histrionic, narcissistic, antisocial), and C (dependent, obsessive-compulsive, avoidant) [12]. The use of the MCMI-III has been decisive because it is highly valid in clinical work with patients. On one hand, it provides us with information about PDs, and on the other, in Axis I, and in Axis II for PDs the MCMI-III gives us scores on the clinical syndromes they present. The MCMI has been used in the evaluation of the AUD as well as other drug use disorder [13,14,15].

Recent research studied the reporting of sex and gender in randomized controlled trials found that 98% of studies described the demographic composition of their participants by sex, however only 6% conducted a subgroup analysis across sex and only 4% reported sex-disaggregated data [16].

Despite that the AUD occur in men and women, sometimes the sex based approached is not considered [17,18,19]. Sex differences in PDs are known. Avoidant, dependent, and paranoid PDs are significantly more frequent in women than men, whereas antisocial PD are more frequent in men compared to women [4]. Moreover, AUD prevalence is higher in men (14.1%) compared to women (5.3%) [20]. However, simultaneous examination of sex differences and personality profiles in AUD remains underexplored.

This study aims to investigate sex differences in personality traits and clusters personality disorders among AUD patients, using the MCMI-III, in a clinical sample requiring treatment.

2. Materials and Methods

2.1. Participants

Patients requesting treatment for alcoholism at the Santa Cruz de Tenerife (Canary Islands, Spain) San Miguel Clinic of Drug Dependence and the Female Alcoholism Unit took part in the study. The sample was recruited between 2016 and 2021. Patients meeting DSM-5 diagnostic criteria for Alcohol Use Disorder (AUD) took part in the study. Patients had a clinical picture between moderate and severe (> 4 symptoms). The selection criteria were age between 18 and 65 years old and fluency in Spanish. Exclusion criteria included physical illness or cognitive impairment. Sex identity was self-reported.

2.2. Personality Assessment

The Millon Clinical Multiaxial Inventory-III (MCMI-III) was used to assess personality. The MCMI-III is a 175-item true-false self-report measure comprising 28 scales, including 10 basic personality pattern scales aligned with DSM-5 PDs [21]. In this research, we will analyse the 10 scales of basic personality patterns corresponding to the DSM-5. According to the standards, values of 60-74 were considered as suggestive of symptoms at a subclinical level, scores between 75-85 were considered as indicative of abnormality, and scores over 85 were considered as presence of abnormality [12]. In this research, we will analyse the dimensional approach to personality (trait scores) as well as the categorical approach to personality (distribution of personality types).

2.3. Study Protocol

Patients requesting treatment for AUD took part in the study. All patients signed an informed consent form and a therapeutic contract after the procedures had been fully explained. Biopsychosocial information was collected through the interdisciplinary clinical history. The therapeutic team (physician, psychologist, and social worker) collected the data. The MCMI-III was administered with the patient in a situation of total alcohol abstinence after detoxification. To monitor patients’ withdrawal status, routine urine drug screenings were performed.

2.4. Statistical Analysis

Continuous variables are presented as means ± SD, categorical variables as frequencies and percentages. Comparisons employed chi-square or Fisher’s exact tests for categorical variables and Student’s t-test for continuous variables. Statistical analyses were performed using the SPSS statistical software package (IBM Corp. Released 2017. IBM SPSS Statistics for Windows, Version 25.0. Armonk, NY: IBM Corp.). All tests were two-tailed, and p values < 0.05 were considered statistically significant.

3. Results

3.1. Sample Characteristics

Two hundred and twenty-three outpatients participated in the study, of whom seven were excluded due to invalid MCMI-III results. The final sample included 216 subjects: 114 women (53%) and 102 men (47%). No participants identified outside the male/female binary.

3.2. Sociodemographic Differences by Sex

Table 1 presents the results of the sociodemographic comparison by sex.

No significant sex differences were observed in age, number of children, employment situation, or educational level. Men were more frequently married or partnered (p = 0.01) while women were divorced/separated more often (p = 0.01). Men lived more frequently with the family of origin (p = 0.04).

3.3. Clinical Characteristics by Sex

Table 2 shows the comparison of clinical characteristics by sex.

Women asked for help significantly before than men, so women had significantly fewer years of consumption than men (p = 0.001). There were no sex differences between women and men regarding the consumption pattern (regular versus occasional) and the type of treatment (Outpatients Addiction Unit, Hospital Detoxification Unit, Residential Unit for Addiction Disorders). Women and men had a high percentage of family with antecedents of drugs consumption, being significantly higher in men than women (p = 0.02). With respect to the secondary use of drugs, men preferred to use more cocaine than women while women preferred to use more benzodiazepines than men (p = 0.001).

3.4. Personality Trait Scores by Sex

Table 3 presents the comparison of the personality trait scores by sex.

Regarding cluster A, not significantly sex differences scores were found. Antisocial, borderline, and narcissistic trait scores did not differ between women and men in cluster B. Women had significantly higher histrionic trait scores than men (p = 0.001). Cluster C dependent (p = 0.002) and obsessive-compulsive (p = 0.001) traits were higher in women. Avoidant trait did not differ between women and men.

3.5. Personality Clusters Trait Type Distribution by Sex

Table 4 presents the comparison of personality traits levels distribution by sex.

No cluster A personality type differences in percentage distribution (paranoid, schizoid and schizotypal) were found by sex. Regarding cluster B type personality distribution by sex, no antisocial, borderline, or histrionic differences by sex were found. Cluster B narcissistic personality disorder was more prevalent in men than women (44% vs. 20%; p = 0.012). In the case of cluster C, no sex differences in avoidant and obsessive-compulsive were found in the personality distribution. However, dependent personality disorder was significantly more frequent in women than men (38% vs. 15%; p = 0.013).

4. Discussion

To our knowledge, this is the first research that study AUD considering the MCMI-III personality clinical inventory, taking into consideration the quantitative, dimensional, approach, (trait score) as well as the percentage levels of personality types, categorial approach (suggestive of symptoms, indicative of abnormality, presence of abnormality,) in a sample of outpatients well balanced by sex. Our findings demonstrate that sex differences vary depending on the analytic approach: while trait scores indicated higher paranoid, histrionic, dependent, and obsessive-compulsive traits in women, only dependent personality disorder prevalence (cluster) and narcissistic personality disorder prevalence in men remained significantly higher when considering categorical thresholds.

Regarding Cluster A, women’s higher paranoid trait scores did not translate into higher cluster prevalence, consistent with recent epidemiological findings [22]. Previous studies report mixed results: some found no sex differences in cluster A traits [23], while other noted greater schizoid personality in men [24].

In Cluster B, women showed elevated histrionic traits dimensionally but not in cluster, aligning with some studies but contrasting with others reporting higher borderline PD in women [25]. Our finding of greater narcissistic personality disorder prevalence in men concurs with contemporary literature [26].

For Cluster C, dependent and obsessive-compulsive traits were higher in women dimensionally; however, only dependent personality disorder prevalence was statistically significant. These results mirror epidemiological trends showing higher dependent PD prevalence in women [27].

Differences in findings between dimensional traits and cluster diagnoses accentuate the importance of integrating both approaches for comprehensive understanding. Dimensional models capture symptoms severity and subclinical variations, while cluster diagnoses facilitate clinical decision-making.

Limitations: The cross-sectional design precludes causal inference. The sample was recruited from a single outpatient unit, limiting generalizability