Cluster analysis of patients with alcohol use disorder featuring alexithymia, depression, and diverse drinking behavior

Abstract

Aim

This study aimed to identify subgroups of alcohol use disorder (AUD) based on a multidimensional combination of alexithymia, depression, and diverse drinking behavior.

Method

We recruited 176 patients with AUD, which were initially divided into non‐alexithymic (n = 130) and alexithymic (n = 46) groups using a cutoff score of 61 on the Toronto Alexithymia Scale (TAS‐20). Subsequently, the profiles of the two groups were compared. Thereafter, a two‐stage cluster analysis using hierarchical and K‐means methods was performed with the Z‐scores from the TAS‐20, the Quick Inventory of Depressive Symptomatology Self‐Report Japanese Version, the 12‐item questionnaire for quantitative assessment of depressive mixed state, and the 20‐item questionnaire for drinking behavior pattern.

Results

In the first analysis, Alexithymic patients with AUD showed greater depressive symptoms and more pathological drinking behavior patterns than those without alexithymia. Cluster analysis featuring alexithymia, depression, and drinking behavior identified three subtypes: Cluster 1 (core AUD type) manifesting pathological drinking behavior highlighting automaticity; Cluster 2 (late‐onset type) showing relatively late‐onset alcohol use and fewer depressive symptoms or pathological drinking behavior; and Cluster 3 (alexithymic type) characterized by alexithymia, depression, and pathological drinking behavior featuring greater coping with negative affect.

Conclusion

The multidimensional model with alexithymia, depression, and diverse drinking behavior provided possible practical classification of AUD. The alexithymic subtype may require more caution, and additional support for negative affect may be necessary due to accompanying mood problems and various maladaptive drinking behaviors.

INTRODUCTION

Alexithymia is a psychological condition characterized by a reduction in or incapacity to experience inner emotions, difficulty in verbalizing emotions, a limited ability to engage in fantasizing, and an absence of a tendency to reflect on one's emotions.¹ This deficit is frequently correlated with unhelpful coping strategies and immature ways of handling stress^2–4 which can worsen depressive symptoms.⁵ Current research focuses on the association between alexithymia and various mental health disorders.⁶ In particular, the effect of alexithymia on alcohol use disorder (AUD) has gained attention in recent years.^7–9

A previous study reported that the prevalence of alexithymia in patients with AUD ranged from 45% to 67%.¹⁰ The co-occurrence rate of alexithymia in the general population is reported to be approximately 6%–10%,^11,12 thereby suggesting that the prevalence of alexithymia in patients with AUD is remarkably high. Alexithymic patients with AUD, who are more vulnerable to stress and heavily impacted by negative emotions,^13,14 are prompted to drink alcohol as a coping strategy.¹⁵ This trait also negatively affects the AUD prognosis.^16,17 Furthermore, this condition is closely associated with anger-driven drinking behavior, which can lead to suicidal thoughts.^18,19 These findings demonstrate a complicated association between alexithymia, negative emotions, and maladaptive drinking behavior in AUD patients.

Meanwhile, AUD-specific drinking behavior consists of diverse patterns, including automatic drinking, which can be aimless or habitual, as well as drinking as a form of coping.^20–22 In particular, automatic drinking behavior is considered one of the core maladaptive behaviors characteristic of AUD, potentially leading to a poorer prognosis with higher relapse rates.²³ However, most previous studies have predominantly examined the relationship between alexithymia and drinking as a coping mechanism, while the relationship with automatic drinking has not been investigated.^15,24,25

Several previous cluster analyses seeking possible AUD subtypes have dealt with the general population, including subthreshold and mild AUD traits,^26,27 although these results may not necessarily be applicable to core AUD, which is encountered in usual clinical practice. Meanwhile, that each patient with AUD has their own combined patterns in terms of alexithymic profiles, mood problems, and maladaptive drinking behavior has also been speculated, thereby suggesting the necessity for individualized treatments. Therefore, cluster analysis using a multidimensional model with alexithymia, mood symptoms, and diverse drinking behavior may be helpful in finding a practical subgrouping of AUD.

Accordingly, we conducted a cluster analysis of alexithymia, depression, and drinking behavior among patients with AUD in the present study. Few comprehensive studies have focused on the associations among these factors or classifications based on a combination of the same factors.

METHODS

Participants

This study included 176 inpatients diagnosed with AUD at the National Hospital Organization Ryukyu Hospital between March 2018 and March 2020. Diagnoses were confirmed by two experienced psychiatrists using DSM-5 criteria.²⁸ The participants consisted of 147 male (83.5%) and 29 female (16.5%) patients, with an overall mean age of 48.7 ± 11.2 years and an age range between 20 and 75 years. Patients with AUD who were unable to understand the questionnaire instructions were excluded from this study.

Measures

The Toronto alexithymia scale (TAS-20)²⁹

The TAS-20 is a gold-standard measure for alexithymia. In the Japanese version of this scale,³⁰ a detailed assessment of the alexithymia subcategories has been provided, which includes difficulty identifying feelings (Cronbach's α = 0.85), difficulty describing feelings (α = 0.72), and externally oriented thinking (α = 0.58). The TAS-20 scores range from 20 to 100. Additionally, previous reports have suggested a cutoff score of 61 on the TAS-20 for distinguishing alexithymia, highlighting its validity.^29,31 Using this cutoff score enables the categorization of patients with pronounced alexithymic traits, thereby enabling a clearer distinction between those with and without significant alexithymia levels.

The quick inventory of depressive symptomatology self-report Japanese version (QIDS-SR-J)³²

The severity of depression was assessed using the QIDS-SR-J, originally developed by Rush et al.³³ This questionnaire was later standardized for Japanese use, with its validity and reliability verified by Fujisawa et al.³² This instrument comprises 16 items, with each item scored from 0 to 3. The overall score is derived by adding the highest score from among the four sleep items (sleep onset insomnia, mid nocturnal insomnia, early morning insomnia, and hypersomnia) (items 1–4); sad mood (item 5); the highest score from the four appetite/weight change items (appetite increase, appetite decrease, weight increase, and weight decrease) (items 6–9); concentration/decision-making (item 10); outlook (self) (item 11); suicidal ideation (item 12); general interest (item 13); energy/fatigability (item 14); and the highest score on the two psychomotor agitation/retardation items (psychomotor slowing and psychomotor agitation) (items 15 and 16). This method yields a total score ranging from 0 to 27.

The 12-item questionnaire for the quantitative assessment of depressive mixed state (DMX-12)³⁴

The DMX-12 was utilized to evaluate the subtle manic and mixed affective components. The psychometric properties, including the reliability and validity of the Japanese version of this scale, have been verified.³⁴ The DMX-12 is a 12-item self-report questionnaire that consists of three subcategories: Spontaneous instability (α = 0.868), which captures symptoms such as restlessness, racing or crowded thoughts, mood lability, inner tension, distractibility, and impulsivity; vulnerable responsiveness (α = 0.826), characterized by hypersensitivity and overreactivity; and disruptive emotion/behavior (α = 0.769), encompassing aggression, irritability, dysphoria, and risk-taking behavior. Items on the DMX-12 are scored on a four-point scale based on symptom frequency over the past week, with options ranging from 0 (never) to 3 (almost always). The total possible score ranges from 0 to 36.

The 20-item questionnaire for drinking behavior patterns (DBP-20)²²

The DBP-20 facilitates the quantitative assessment of diverse drinking behavior intensities in patients with AUD, encompassing coping drinking to manage negative emotions, automatic drinking including habitual patterns, enhancement drinking aimed at mood elevation, and social drinking to engage in social situations (coping with negative affect (α = 0.913), automaticity (α = 0.868), enhancement (α = 0.719), and social use (α = 0.687), respectively). This scale, developed as a Japanese version, has undergone verification of its validity and reliability.²² Scoring for each DBP-20 item is conducted on a four-point scale reflecting the frequency of specific drinking behavior patterns, with choices ranging from 0 (never) to 3 (almost always). The scores for each subgroup are calculated as follows: coping with negative affect (7 items) 0–21, automaticity (8 items) 0–24, enhancement (3 items) 0–9, and social use (2 items) 0–6.

The alcohol use disorders identification test (AUDIT)³⁵

The AUDIT, translated globally, is recognized as a gold standard 10-item self-report questionnaire, with its validity confirmed in the Japanese context.³⁶ This scale facilitates the evaluation of risky or hazardous drinking patterns and the detection of alcohol-related issues, providing a total score spectrum from 0 to 40.

Variables

At the time of admission, the patients' biological and sociological information was meticulously gathered. These included age, sex, total years of education, current employment status, smoking status, marital status, living arrangements (specifically, living alone), psychiatric comorbidities, the age when initiating alcohol use, and pharmacotherapy (antidepressants, antipsychotics, hypnotics, anxiolytics, and antiseizure drugs). The questionnaires TAS-20, AUDIT, QIDS-SR-J, DMX-12, and DBP-20 were administered within the first week of admission, after the influence of alcohol had dissipated and alcohol withdrawal symptoms were no longer present.

Statistical analyses

In this study, the TAS-20 with a cutoff score of 61 was used to classify participants into non-alexithymic and alexithymic patients with AUD. Demographic variables were compared between the two groups by conducting Student's t-test or Mann–Whitney U-test for continuous variables and Chi-squared test or Fisher's Exact Test for categorical variables.

A two-stage cluster analysis was conducted. The cluster variables included QIDS-SR-J, DMX-12, TAS-20, and DBP-20 subscale scores (coping with negative affect, automaticity, enhancement, and social use, respectively). First, the raw scores of the cluster variables were standardized and transformed into Z-scores. Hierarchical cluster analysis using Ward's method³⁷ was conducted to estimate the appropriate cluster number. We used the squared Euclidean distance³⁸ as the dissimilarity measure. A dendrogram plot was used to inspect the overall structure of the data and determine the optimal cluster solution. The elbow method³⁹ was used to confirm the optimal cluster number. In the second stage, K-means clustering⁴⁰ was performed using the predefined optimal number of clusters.

Following the two-stage cluster analysis, a one-way analysis of variance (ANOVA) was conducted to identify significant differences across the three clusters. For continuous variables, Tukey's Honestly Significant Difference (HSD) test was employed in post-hoc analyses to pinpoint specific differences between clusters. For categorical variables, Chi-squared tests were conducted to evaluate differences among the clusters.

A two-tailed p-value less than 0.05 was considered statistically significant. All statistical analyses were performed using SPSS 28.0.0.0J for Windows (SPSS Japan, Tokyo, Japan).

Ethics procedures

Ethical considerations were strictly adhered to throughout the study. Each participant voluntarily signed an informed consent form prior to participation. To ensure confidentiality, all participant data were anonymized and analyzed in an aggregate form. The participants were fully briefed on the objectives of the study, confidentiality measures, and their freedom to discontinue participation at any time. The study methodology was approved by the Ethics Committee of the National Hospital Organization Ryukyu Hospital under the approval code 29–21.

RESULTS

A section of the study participants presented with psychiatric comorbidities, including depressive disorder in 17 patients, bipolar disorder in six, anxiety disorder in three, panic disorder in four, and alcohol-induced psychosis in two. All the participants underwent the same standardized alcohol rehabilitation program that primarily included cognitive-behavioral therapy, motivational interviewing, and group therapy sessions. The participants' medications during their hospitalization included antidepressants (paroxetine: 3; escitalopram: 2; sertraline: 1; mirtazapine: 3; trazodone: 5; duloxetine: 2), antipsychotics (risperidone: 4; quetiapine: 17; olanzapine: 3; perospirone: 1; chlorpromazine: 1; levomepromazine: 1), hypnotics (brotizolam: 9; flunitrazepam: 6; nitrazepam: 2; zolpidem: 5; zopiclone: 3; eszopiclone: 8; suvorexant: 11; ramelteon: 5), benzodiazepine anxiolytics (alprazolam: 1; lorazepam: 1; bromazepam: 2; etizolam: 2; diazepam: 1), and antiseizure medications (sodium valproate: 3; carbamazepine: 1; lamotrigine: 2; levetiracetam: 2; gabapentin: 1; perampanel hydrate 1).

Table 1 presents the comparison of demographics and clinical characteristics between non-alexithymic (n = 130) and alexithymic (n = 46) patients with AUD. The AUD patients with alexithymia exhibited a shorter education period, lower employment rates, a higher prevalence of smoking, and higher rates of antidepressant and anxiolytic use than those without alexithymia. Alexithymic patients exhibited higher QIDS-SR-J and DMX-12 scores as well as higher scores of maladaptive drinking behavior patterns (coping with negative affect, automaticity, and enhancement) of the DBP-20 than non-alexithymic patients.

TABLE 1 Comparison of demographics and clinical characteristics between non-alexithymic and alexithymic patients with alcohol use disorder.

Variable	Non-Alexithymic AUD patients (n = 130)	Alexithymic AUD patients (n = 46)	p-Value
Age (years)a	49.4 ± 10.9	46.8 ± 12.0	0.181
Sexc
Male (%)	111 (85.4)	36 (78.3)
Female (%)	19 (14.6)	10 (21.7)	0.263
Education (years)b	11.9 ± 2.1	11.2 ± 1.6	0.037
Employed (%)c	62 (47.7)	12 (26.1)	0.011
Marital statusc
Married (%)	47 (36.2)	15 (32.6)
Divorced, unmarried, or widowed (%)	83 (63.8)	31 (67.4)	0.665
Current smoker (%)c	94 (72.3)	40 (87.0)	0.045
Living alone (%)c	39 (30.0)	16 (34.8)	0.548
Psychiatric comorbidities (%)c	24 (27.5)	8 (23.5)	0.872
Age at first alcohol use (years)b	16.5 ± 3.2	15.7 ± 4.7	0.052
Antidepressants (%)c	6 (4.6)	10 (21.7)	0.001
Antipsychotics (%)c	18 (13.8)	9 (19.6)	0.355
Hypnotics (%)c	31 (23.8)	18 (39.1)	0.047
Anxiolytics (%)d	2 (1.5)	5 (10.9)	0.005
Antiseizure drugs (%)d	7 (5.4)	3 (6.5)	0.722
AUDIT total scorea	27.5 ± 6.3	28.4 ± 7.5	0.311
QIDS-SR-J total scoreb	9.1 ± 4.4	13.9 ± 4.8	<0.001
DMX-12 total scoreb	8.3 ± 5.7	14.6 ± 7.8	<0.001
DBP-20 subscale scoresb
Coping with negative affect	10.0 ± 5.8	14.3 ± 5.5	<0.001
Automaticity	13.2 ± 6.3	16.4 ± 4.9	0.004
Enhancement	2.8 ± 2.3	3.9 ± 2.4	0.007
Social use	3.2 ± 1.8	2.9 ± 1.7	0.293

Cluster analysis identified three distinct subgroups of patients with AUD. Figure 1 displays the unique profiles of these clusters, based on Z-scores from the TAS-20, QIDS-SR-J, DMX-12, and DBP-20. Specifically focusing on the TAS-20, QIDS-SR-J, DMX-12, and the coping with negative affect subscale of the DBP-20, the three clusters displayed similar patterns: Cluster 1 (C1) displayed intermediate values, Cluster 2 (C2) had the lowest values, and Cluster 3 (C3) exhibited the highest values. Regarding other DBP-20 subscales, the Z-scores for automaticity were higher in C1 (Z-score = 0.6 ± 0.7, p < 0.001) and C3 (Z-score = 0.4 ± 0.8, p < 0.001) than in C2 (Z-score = −0.8 ± 0.8). Similarly, C1 (Z-score = 0.2 ± 1.0, p < 0.001) and C3 (Z-score = 0.6 ± 0.9, p < 0.001) had higher Z-scores for enhancement than C2 (Z-score = −0.6 ± 0.7). For the social use subscale, C1 (Z-score = 0.5 ± 1.0) showed higher Z-scores than C2 (Z-score = −0.5 ± 0.9, p < 0.001) or C3 (Z-score = −0.1 ± 0.9, p = 0.004).

[IMAGE OMITTED. SEE PDF]

Table 2 summarizes the profiles of three distinct clusters examined by the one-way ANOVA or Chi-squared test, highlighting significant differences in age, sex, psychiatric comorbidities, age at first alcohol use, total TAS scores, total QIDS-SR-J scores, total DMX-12 scores, and DBP-20 subscale scores. The post-hoc analyses revealed partial differences or significant gradations in various characteristic profiles among the three clusters. Considering the demographic data and psychiatric manifestations (Table 2), the patients with AUD were classified into the following three subtypes, that is, C1: predominantly featuring maladaptive drinking behavior highlighting automaticity (core AUD type); C2: featuring relatively late-onset alcohol use and fewer depressive symptoms or pathological drinking (late-onset type); and C3: characterized by alexithymia, depression, and maladaptive drinking behavior more featuring coping with negative affect (alexithymic type).

TABLE 2 Demographics and clinical characteristics of patients with alcohol use disorder across three clusters.

	C1 (n = 66)	C2 (n = 70)	C3 (n = 40)	p-Value	Comparisons among clusters	C1 vs. C2 (p-value)	C2 vs. C3 (p-value)	C1 vs. C3 (p-value)
Age (years)	47.3 ± 9.7	51.6 ± 12.2	46.2 ± 11.0	0.019	C3 < C2, C3 ≈ C1, C1 ≈ C2	–	0.036	–
Sex (male) (%)	58 (87.9)	63 (90.0)	26 (65.0)	0.001	C3 < C1 ≈ C2	–	0.001	0.005
Education (years)	12.0 ± 2.2	11.8 ± 3.2	11.2 ± 1.8	–	–	–	–	–
Married (%)	24 (38.6)	27 (38.6)	11 (27.5)	–	–	–	–	–
Employed (%)	28 (42.4)	34 (48.6)	12 (30.0)	–	–	–	–	–
Current smoker (%)	49 (74.2)	53 (75.7)	32 (80.0)	–	–	–	–	–
Living alone (%)	22 (33.3)	20 (28.6)	13 (32.5)	–	–	–	–	–
Psychiatric comorbidities (%)	12 (18.2)	7 (10.0)	13 (32.5)	0.013	C2 < C3 C2 ≈ C1, C1 ≈ C3	–	0.003	–
Depressive disorder (%)	6 (9.1)	4 (5.7)	7 (17.5)	–	C2 < C3 C2 ≈ C1, C1 ≈ C3	–	0.047	–
Bipolar disorder (%)	4 (6.1)	0 (0.0)	2 (5.0)	–	–	–	–	–
Anxiety disorder (%)	0 (0.0)	2 (2.9)	1 (2.5)	–	–	–	–	–
Panic disorder (%)	2 (3.0)	0 (0.0)	2 (5.0)	–	–	–	–	–
Alcohol-induced psychosis (%)	0 (0.0)	1 (1.4)	1 (2.5)	–	–	–	–	–
Age at first alcohol use (years)	15.4 ± 3.1	17.5 ± 3.2	15.7 ± 4.8	0.002	C1 ≈ C3 < C2	0.002	0.031	–
Antidepressants (%)	5 (7.6)	1 (1.4)	10 (25.0)	<0.001	C1 ≈ C2 < C3	–	<0.001	0.020
Antipsychotics (%)	10 (15.2)	5 (7.1)	12 (30.0)	0.006	C2 < C3 ≈ C1	–	0.002	–
Hypnotics (%)	21 (31.8)	13 (18.6)	40 (37.5)	–	C2 < C3 ≈ C1	–	0.040	–
Anxiolytics (%)	1 (1.5)	1 (1.4)	5 (12.5)	0.007	C1 ≈ C2 < C3	–	0.023	0.028
Antiseizure drugs (%)	3 (4.5)	3 (4.3)	4 (10.0)	0.405	–	–	–	–
AUDIT total score	27.8 ± 5.7	25.7 ± 6.7	29.6 ± 7.5	–	C2 < C3, C2 ≈ C1, C1 ≈ C3	–	0.045	–
TAS-20 total score	51.5 ± 7.1	49.1 ± 8.0	65.9 ± 5.9	<0.001	C1 ≈ C2 < C3	–	<0.001	<0.001
QIDS-SR-J total score	9.6 ± 4.3	8.0 ± 3.9	15.6 ± 3.7	<0.001	C2 < C1 < C3	0.041	<0.001	<0.001
DMX-12 total score	9.5 ± 5.3	5.7 ± 4.2	18.2 ± 5.8	<0.001	C2 < C1 < C3	<0.001	<0.001	<0.001
DBP-20 subscale scores
Coping with negative affect	13.4 ± 4.6	6.2 ± 4.2	16.1 ± 4.2	<0.001	C2 < C1 < C3	<0.001	<0.001	0.006
Automaticity	18.0 ± 4.0	9.0 ± 4.8	16.5 ± 4.9	<0.001	C2 < C1 ≈ C3	<0.001	<0.001	–
Enhancement	3.8 ± 2.4	1.8 ± 1.6	4.4 ± 2.1	<0.001	C2 < C1 ≈ C3	<0.001	<0.001	–
Social use	4.0 ± 1.7	2.3 ± 1.5	3.0 ± 1.6	<0.001	C2 ≈ C3 < C1	<0.001	–	0.004

DISCUSSION

In our study, the prevalence of alexithymia among patients with AUD (26.1%) was lower than the 45–67% previously reported in the AUD population.¹⁰ This difference might be at least partly explained by our sample's male-dominant sex distribution (83.5%) since the previous research suggested a higher co-occurrence of alexithymia among female patients with AUD.⁴¹

The present study also demonstrated that alexithymic patients with AUD had poorer social backgrounds, such as lower educational levels and higher unemployment rates, than those without alexithymia, which is consistent with a previous study exploring the association between alexithymia and low socioeconomic status in the general population.⁴² Although the causal relationship between alexithymia and poor socioeconomic background in patients with AUD is unknown, alexithymic patients with AUD may have more difficulty in problem-solving due to their poorer capability for social adjustment than those without alexithymia. In addition, undifferentiated discomfort derived from interpersonal conflict or social maladaptation can lead to self-medicated substance use as a stereotypical measure to escape reality and mask psychological stress, which is likely to result in unhealthy mental outcomes in these patients.

Furthermore, this study also demonstrated that AUD patients with alexithymia experience higher levels of depressive/mixed symptoms and exhibit a broader spectrum of maladaptive drinking behavior, including coping with negative affect, automaticity, and enhancement (Table 1). Although previous studies have only focused on pathological drinking based on aspects of negative coping style in alexithymic patients with AUD,^13–15 our findings suggest that alexithymia is also accompanied by other maladaptive drinking patterns such as automatic and motivation-enhancing drinking. More frequent mood problems and broader patterns of maladaptive drinking linked with alexithymia appear to be the potential targets of pharmacological and behavioral interventions, respectively. Accordingly, considering the combined features of alexithymia, depression, and pathological drinking, rather than dichotomous grouping solely by alexithymia with the arbitrary TAS-20 cut-off, may shed new light on individualized AUD treatment strategies. Therefore, we employed cluster analysis featuring alexithymia, depression, and drinking behavior for a multidimensional classification of AUD.

The present study provided a three-cluster model of AUD based on Z-scores from the total scores of the TAS-20, QIDS-SR-J, and DMX-12, as well as the subscale scores of the DBP-20 (Table 2 and Figure 1). C1 is characterized by mild depression without alexithymia but predominantly features a variety of pathological drinking behavior patterns, with automaticity, the core AUD type, featured most (Table 2 and Figure 1). Our previous research indicated that automatic drinking behavior is the most representative drinking behavior of AUD and differentiates individuals with AUD from non-alcoholic individuals,²² which can lead to an increased risk of future relapse of alcohol use.²³ Considering these findings, C1 may be regarded as the core AUD subgroup that engages in automatic drinking, easily leading to habitual alcohol use, and is more likely to relapse.

C2 generally exhibited low scores on the TAS-20, QIDS-SR-J, DMX-12, AUDIT, and DBP-20 (Table 2 and Figure 1). Furthermore, this cluster consisted of older individuals with lower depression or other psychiatric comorbidities and displayed relatively late-onset alcohol use (Table 2). Thus, C2 is manifested by relatively late-onset AUD with fewer depressive symptoms or pathological drinking patterns (late-onset type), which is similar to the milder form of AUD with late onset and lower severity identified in a previous study.²⁶ In particular, the lower automatic drinking than other subgroups (Table 2) may lead to a lower relapse risk of alcohol use in this subgroup, according to our previous report that demonstrated a close association between automatic drinking and future relapse of alcohol use.²³

The present study specifically highlighted the C3 subgroup, which is characterized by alexithymia, depression, and pathological drinking behavior, most featuring greater coping with negative affect (alexithymic type). This group consisted of younger individuals and more females, with a higher prevalence of psychiatric comorbidities, particularly including depressive disorders, and greater use of antidepressants, antipsychotics, hypnotics, and anxiolytics (Table 2). The coexistence of alexithymia and high levels of dysphoric depression in the C3 group is consistent with a previous report showing a close association between alexithymia and psychiatric disorders.⁶ Meanwhile, this alexithymic cluster, which included more young and female inpatients with AUD, may be a unique subgroup, considering that older age and male sex have been associated with alexithymia in community respondents⁴² and outpatients with AUD.⁴³

Although plausible personalized approaches based on our preliminary clustering of AUD are still speculative and need to be further confirmed by verified evidence in the future, some clinical implications might be drawn from a therapeutic perspective. In this study, the alexithymic subgroup (C3) may overuse alcohol for both negative and positive coping to mask depressed moods and enhance their motive for action. Meanwhile, our recent study demonstrated that patients with AUD who habitually drink to cope with negative affect experienced more residual depression even after alcohol detoxification.⁴⁴ Therefore, such patients may need intensive psychoeducation to understand the complicated interactions among alexithymia, depression, and drinking behavior, and additional support for negative affect may be necessary. In contrast, C1 is predominantly characterized by automatic drinking, which is the core behavior differentiating AUD patients from normal drinkers and has been suggested as a potential marker for future relapse of alcohol use.²³ Due to this subgroup's core feature of addictive behavior, a standardized alcohol rehabilitation program would be the most essential.

In conclusion, the multidimensional model with alexithymia, depression, and diverse drinking behavior provided a possible practical classification of AUD. The alexithymic subtype may require more caution, and additional support for negative affect may be necessary due to accompanying mood problems and various maladaptive drinking behaviors. However, treatment approaches specific to each subgroup need to be validated in future studies.

This study had several limitations. First, the findings of the cluster analysis were derived from a relatively small group of inpatients with chronic AUD, predominantly male and middle-aged individuals. Further research with a more substantial and diverse sample is necessary to ensure the accuracy and generalizability of the findings to the broader AUD population. Second, the psychological evaluations employed in this study may have been influenced by self-report biases, including psychological denial and underestimation. Third, numerous models and methods exist for conducting cluster analyses in AUD research. Thus, further studies are essential to conduct appropriate analyses that are best suited for patients with AUD.

AUTHOR CONTRIBUTIONS

Kazuhiro Kurihara, Hiroyuki Enoki, Yoshikazu Takaesu, and Tsuyoshi Kondo designed the study, wrote the protocol, performed statistical analyses, and wrote the manuscript. Kazuhiro Kurihara, Hiroyuki Enoki, Hotaka Shinzato and Tsuyoshi Kondo wrote and verified the manuscript. Tsuyoshi Kondo raised funding. All authors contributed to the manuscript and approved this submission.

FUNDING INFORMATION

This study was supported by the JSPS KAKENHI (grant numbers JP17K10311 and JP21K07504). The funding source had no role in the design, practice, or analysis of this study.

CONFLICT OF INTEREST STATEMENT

The authors declare no conflict of interest.

DATA AVAILABILITY STATEMENT

The data from this study have not been made publicly available because the disclosure of individual data was not included in the research protocol, and consent for public data sharing was not obtained from the participants.

ETHICS STATEMENT

Approval of the Research Protocol by an Institutional Reviewer Board: The Ethics Committee of the National Hospital Organization Ryukyu Hospital approved the study protocol (approval number 29–21).

Informed Consent: All participants provided written informed consent before participating in the study.

Registry and the Registration No. of the Study/Trial: n/a.

Animal Studies: n/a.

References

Larsen JK, Brand N, Bermond B, Hijman R. Cognitive and emotional characteristics of alexithymia: a review of neurobiological studies. J Psychosom Res. 2003;54(6):533–541.

Taurino A, Antonucci LA, Taurisano P, Laera D. Investigating defensive functioning and alexithymia in substance use disorder patients. BMC Psychiatry. 2021;21(1):337.

Myers LB, Derakshan N. The relationship between two types of impaired emotion processing: repressive coping and alexithymia. Front Psychol. 2015;6:809.

Joybari MT. Depression and interpersonal problems in adolescents: their relationship with alexithymia and coping styles. Iran J Psychiatry Behav Sci. 2014;8(4):38–45.

Honkalampi K, Koivumaa‐Honkanen H, Lehto SM, Hintikka J, Haatainen K, Rissanen T, et al. Is alexithymia a risk factor for major depression, personality disorder, or alcohol use disorders? A prospective population‐based study. J Psychosom Res. 2010;68(3):269–273.

Morie KP, Crowley MJ, Mayes LC, Potenza MN. The process of emotion identification: considerations for psychiatric disorders. J Psychiatr Res. 2022;148:264–274.

Loas G, Otmani O, Lecercle C, Jouvent R. Relationships between the emotional and cognitive components of alexithymia and dependency in alcoholics. Psychiatry Res. 2000;96(1):63–74.

Kopera M, Trucco EM, Jakubczyk A, Suszek H, Michalska A, Majewska A, et al. Interpersonal and intrapersonal emotional processes in individuals treated for alcohol use disorder and non‐addicted healthy individuals. Addict Behav. 2018;79:8–13.

Youssef L, Salameh P, Sacre H, Akel M, Hallit S, Obeid S. Association between alexithymia, emotional intelligence, smoking addiction, and alcohol use disorder among a sample of Lebanese adults. PLoS One. 2023;18(11): [eLocator: e0295114].

Thorberg FA, Young RMD, Sullivan KA, Lyvers M. Alexithymia and alcohol use disorders: a critical review. Addict Behav. 2009;34(3):237–245.

Kokkonen P, Karvonen JT, Veijola J, Läksy K, Jokelainen J, Järvelin MR, et al. Prevalence and sociodemographic correlates of alexithymia in a population sample of young adults. Compr Psychiatry. 2001;42(6):471–476.

Hintikka J, Honkalampi K, Lehtonen J, Viinamäki H. Are alexithymia and depression distinct or overlapping constructs?: a study in a general population. Compr Psychiatry. 2001;42(3):234–239.

Evren C, Cagil D, Ulku M, Ozcetinkaya S, Gokalp P, Cetin T, et al. Relationship between defense styles, alexithymia, and personality in alcohol‐dependent inpatients. Compr Psychiatry. 2012;53(6):860–867.

Coriale G, Messina MP, Geticarelli S, Battagliese G, Delfino D, Fiorentino D, et al. Physiological responses to induced stress in individuals affected by alcohol use disorder with dual diagnosis and alexithymia. Clin Ter. 2020;171(2):e120–e129.

Linn BK, Zhao J, Bradizza CM, Lucke JF, Ruszczyk MU, Stasiewicz PR. Alexithymia disrupts emotion regulation processes and is associated with greater negative affect and alcohol problems. J Clin Psychol. 2021;77(12):2915–2928.

Loas G, Fremaux D, Otmani O, Lecercle C, Delahousse J. Is alexithymia a negative factor for maintaining abstinence? A follow‐up study. Compr Psychiatry. 1997;38(5):296–299.

Maillard A, Laniepce A, Segobin S, Lahbairi N, Boudehent C, Vabret F, et al. Prognostic factors for low‐risk drinking and relapse in alcohol use disorder: a multimodal analysis. Addict Biol. 2022;27(6): [eLocator: e13243].

Ghorbani F, Khosravani V, Sharifi Bastan F, Jamaati AR. The alexithymia, emotion regulation, emotion regulation difficulties, positive and negative affects, and suicidal risk in alcohol‐dependent outpatients. Psychiatry Res. 2017;252:223–230.

Kealy D, Ogrodniczuk JS, Rice SM, Oliffe JL. Alexithymia, suicidal ideation and health‐risk behaviours: a survey of Canadian men. Int J Psychiatry Clin Pract. 2018;22(1):77–79.

Koob GF. A role for brain stress systems in addiction. Neuron. 2008;59(1):11–34.

Everitt BJ, Robbins TW. From the ventral to the dorsal striatum: devolving views of their roles in drug addiction. Neurosci Biobehav Rev. 2013;37(9):1946–1954.

Kurihara K, Shinzato H, Koda M, Enoki H, Otsuru T, Takaesu Y, et al. Development of a 20‐item questionnaire for drinking behavior pattern (DBP‐20) toward personalized behavioral approaches for alcohol use disorder. Alcohol. 2022;101:9–16.

Kurihara K, Shinzato H, Takaesu Y, Kondo T. Associations between relapse and drinking behaviors in patients with alcohol use disorders: a 6‐month prospective study. Neuropsychopharmacol Rep. 2023;43(4):633–640.

Luminet O, de Sousa C, Uva M, Fantini C, de Timary P. The association between depression and craving in alcohol dependency is moderated by gender and by alexithymia factors. Psychiatry Res. 2016;239:28–38.

Thorberg FA, Young RMD, Hasking P, Lyvers M, Connor JP, London ED, et al. Alexithymia and alcohol dependence: the roles of negative mood and alcohol craving. Subst Use Misuse. 2019;54(14):2380–2386.

Moss HB, Chen CM, Yi HY. Subtypes of alcohol dependence in a nationally representative sample. Drug Alcohol Depend. 2007;91(2–3):149–158.

Müller M, Ajdacic‐Gross V, Vetrella AB, Preisig M, Castelao E, Lasserre A, et al. Subtypes of alcohol use disorder in the general population: a latent class analysis. Psychiatry Res. 2020;285: [eLocator: 112712].

American Psychiatric Association. Diagnostic and statistical manual of mental disorders. 5th ed. Washington, DC: American Psychiatric Association; 2013.

Taylor GJ, Bagby RM, Parker JDA. The 20‐item Toronto alexithymia scale: IV. Reliability and factorial validity in different languages and cultures. J Psychosom Res. 2003;55(3):277–283.

Gen K, Motonari M, Atsuyuki A, Akinori N, Haruo S, Ichiko O, et al. The reliability and factorial validity of the Japanese version of the 20‐item Toronto alexithymia scale (TAS‐20). Jpn J Psychosom Med. 2003;43:839–846.

Taylor GJ, Bagby RM, Parker JD. Disorders of affect regulation: alexithymia in medical and psychiatric illness. Cambridge, UK: Cambridge University Press; 1999.

Fujisawa D, Nakagawa A, Tajima M, Sado M, Kikuchi T, Iba M, et al. Cross_cultural adaptation of the quick inventory of depressive symptomatology‐self report (QIDS‐SR‐J) Japanese. J Stress Sci. 2008;25(1):43–52.

Rush AJ, Trivedi MH, Ibrahim HM, Carmody TJ, Arnow B, Klein DN, et al. The 16‐item quick inventory of depressive symptomatology (QIDS), clinician rating (QIDS‐C), and self‐report (QIDS‐SR): a psychometric evaluation in patients with chronic major depression. Biol Psychiatry. 2003;54(5):573–583.

Shinzato H, Koda M, Nakamura A, Kondo T. Development of the 12‐item questionnaire for quantitative assessment of depressive mixed state (DMX‐12). Neuropsychiatr Dis Treat. 2019;15:1983–1991.

Saunders J, Aasland O, Babor T, de la Fuente J, Grant M. Development of the alcohol use disorders identification test (AUDIT): WHO collaborative project on early detection of persons with harmful alcohol consumption‐II. Addiction. 1993;88(6):791–804.

Hiro H, Shima S. Availability of the alcohol use disorders identification test (AUDIT) for a complete health examination in Japan. Japanese J Alcohol Stud Drug Depend. 1996;31(5):437–450.

Ward JH. Hierarchical grouping to optimize an objective function. J Am Stat Assoc. 1963;58(301):236–244.

Everitt BS, Landau S, Leese M. Cluster analysis. 4th ed. London: Arnold; 2001.

Hastie T, Tibshirani R, Friedman J. The elements of statistical learning: data mining, inference, and prediction. 2nd ed. New York, NY: Springer; 2009.

Jain AK. Data clustering: 50 years beyond K‐means. Pattern Recogn Lett. 2010;31(8):651–666.

Cruise KE, Becerra R. Alexithymia and problematic alcohol use: a critical update. Addict Behav. 2018;77:232–246.

Lane RD, Sechrest L, Riedel R. Sociodemographic correlates of alexithymia. Compr Psychiatry. 1998;39(6):377–385.

Thorberg FA, Hasking P, Huang YL, Lyvers M, Young RMD, Connor JP, et al. The influence of alexithymia on alcohol craving, health‐related quality of life and gender in alcohol‐dependent outpatients. J Psychoactive Drugs. 2020;52(4):366–376.

Kurihara K, Shinzato H, Takaesu Y, Kondo T. Drinking behavior patterns may be associated with persistent depressive symptoms after alcohol abstinence in alcohol use disorder. Neuropsychopharmacol Rep. 2024. [DOI: https://dx.doi.org/10.1002/npr2.12429]. Epub ahead of print.