Personality traits and personality problems in Korsakoff syndrome patients

ABSTRACT

Personality research is of relevance because it provides insights into the psychological strengths and vulnerabilities of a person. Korsakoff’s syndrome (KS) is a severe neuropsychiatric disorder following acute thiamine deficiency, usually as a consequence of alcohol-use disorders. Research on personality traits of KS patients is currently limited. The aim of the current exploratory study was therefore to examine the personality traits and problems of KS patients. We assessed self-reported and informant-reported personality traits and problems in institutionalized KS patients (n = 30). Results indicate that the psychotic personality organization was more frequently present in KS patients compared to a psychiatric reference population, leading to increased vulnerability for the development of severe psychiatric issues. Informants observed more premorbid and current cluster B (“emotional”) and C (“anxious”) personality behavior in comparison to the general population. Also, rigid- and socially avoidant behavior in KS patients was observed to increase over the years. They also observed personality problems such as emotional-, unpredictable-, dramatic-, anxious and fearful behavior in the present and in the past. KS patients reported themselves as healthy individuals, indicating a lack of self-awareness. We recommend training programs for the medical team focussed on psycho-education and supportive interventions for patients with complex personality problems concomitant to KS.

KEYWORDS:

• Personality
• Korsakoff syndrome
• personality organization
• self-report
• informant-report

Introduction

Korsakoff syndrome (KS) is a chronic neuropsychiatric syndrome as a consequence of a shortage of thiamine (vitamin B1; Isenberg-Grzeda et al., 2012). A major cause of thiamine deficiency is chronic alcohol abuse with concomitant malnutrition. Cognitive disorders are a hallmark sign of KS. The most striking symptom of KS is memory dysfunction, in which remote and recent memory are affected. Especially the ability to recall the occurrence of events is often impaired (Kopelman et al., 2009). Another characteristic of KS is executive dysfunction, such as poor judgment, difficulties in planning abilities, and impairments in working memory and cognitive flexibility (Moerman van Den Brink et al., 2019).

KS patients often display interpersonal problems due to the aforementioned cognitive impairments. Psychological risk factors that come along with the syndrome are a further cause for the problems KS patients frequently experience in relationships. For instance, KS patients often display a diminished theory of mind (Drost et al., 2018) and impaired emotional decoding functions (Montagne et al., 2006). These two aspects contribute to severe errors in the interpretation of others, leading to impairments in interpersonal communication. Dysfunctional communication might be one of the reasons that KS patients tend to live in social isolation and report feelings of severe loneliness (Oudman et al., 2018). Other neuropsychiatric and behavioral consequences of KS that are frequently reported by nursing staff are depression, aggression, irritability, apathy, and disinhibition of behavior (Gerridzen et al., 2018).

As aforementioned, KS patients frequently display serious difficulties in relating to others, and other neuropsychiatric symptoms. However, it is currently unknown whether these symptoms in KS are an expression of the disorder itself or rather caused by a pattern of risk factors ultimately leading to KS. For instance, eating disorders, psychiatric disorders and chemotherapy are indicated to increase the chance of developing KS later in life (Isenberg-Grzeda et al., 2016; Nikolakaros et al., 2016). Another risk factor might be adverse personality development of an individual. Personality can be defined as the differences between individuals in the form of typical ways of thinking, feeling, and behaving that are formed by environmental and biological aspects. Personality traits are often relatively stable over time (Kazdin, 2000). Distinguishing personality traits allows us to group individuals according to their behavior to better predict future behavior (Skóra et al., 2020). Research on personality traits in several psychiatric disorders indicated that knowledge about personality traits adds to the complexity of diagnosis, prevention, and treatment of a mental disorder. Personality is also a main contributing factor to pathogenesis because it influences social functioning, cognition and symptoms (Compton et al., 1988; Gurrera et al., 2014). For instance, various studies showed that personality features predict the start and course of schizophrenia (Gleeson et al., 2005; Lönnqvist et al., 2009; van Os & Jones, 2001). Furthermore, knowledge about personality can help in the formation of realistic expectations for change, matching interventions to personality, addressing change efforts, and investigating the opportunities to increase self-awareness (Harkness & Lilienfeld, 1996).

One of the determinants that can cause severe problems later in life is vulnerable personality development (Kernberg, 1984). As shown in Figure 1, genetic factors and adverse childhood events, such as child maltreatment, put people at risk for developing a vulnerable personality organization early in life. The presence of a vulnerable personality organization is suggested to be stable and is associated with pathological personality traits, such as emotional instability or impulsivity, which increases the risk of developing personality disorders later in life (L. Eurelings-Bontekoe & Snellen, 2021; Kernberg, 1984). However, behavioral- and personality changes have also been associated as a consequence of neuropsychiatric disorders in earlier studies (Miyoshi et al., 2010). This suggests that KS patients may either have a vulnerable personality profile, or could also suggest that KS has detrimental effects on the personality development of KS patients.

Figure 1. Schematic representation of the development of a vulnerable personality organization to a personality disorder.

Although it is known that KS patients display a variety of neuropsychiatric symptoms, personality traits of KS patients have not yet been systematically studied. The relevance of examining KS and personality is that it might provide more clarity about the psychological vulnerability of those at risk of developing KS. Furthermore, personality assessment can add to the direction of the treatment of KS patients and help generate intervention strategies (van Alphen et al., 2012). Egger et al. (2002), and also S. J. Walvoort et al. (2016) found that KS patients form a distinctive group regarding the results on the MMPI-2. KS patients display an extraordinary flat profile around the normative mean of 50, which is remarkable because it suggests that the patients display no psychiatric issues. In patients with alcohol use disorders without cognitive deficits, Fehrman et al. (2017) found high levels of neuroticism, impulsivity, and sensation-seeking in a large database of drug- and alcohol-dependent patients. Other studies found increased aggressiveness, psychoticism, negative affect, depression, and anxiety, in severe alcohol use disorders (Fromme et al., 2004; Tragesser et al., 2007; Trull et al., 2004). In contrast, Mulder (2002) suggested that there is not a specific relationship between alcohol use disorders and personality traits, and that only a small proportion of the risk for alcohol use disorders is explained by personality traits. In light of this lack of consensus, it is relevant to investigate whether similar personality traits exist in KS patients and to examine if these traits were already present in the premorbid state or if these traits have been developed due to KS.

The main aim of this study was to examine if there is a typical KS personality profile. We examined if certain premorbid personality traits and organizations can put people at risk to develop KS later in life and we also examined if certain current personality traits have been developed due to KS. Therefore, at first, we investigated if there were typical KS (premorbid) personality organizations and/or self-reported KS personality traits. The second aim was to investigate if there were typical informant-reported KS personality traits and if there were differences in informant-reported current- and premorbid personality traits due to KS. The third goals was to examine if there was a relationship between current- and premorbid personality traits. We expected a positive relationship between current- and premorbid personality traits because research showed that personality traits are relatively stable over time (Costa & McCrae, 1988). The last goal was to see if KS patients showed more signs of personality disorders compared to what is typically seen in the general population, given that these disorders can only be diagnosed when there’s no physical cause for the symptoms. Based on the observations that KS patients display more behavioral problems, we expected that the prevalence rate of personality disorders characteristics is higher in KS patients compared to the general population.

Methods

In- and exclusion criteria

The inclusion criteria were that the DSM-5 criteria for the alcohol-induced major neurocognitive disorder (amnestic-confabulatory type) had to be fulfilled by all patients (American Psychiatric Association, 2013). KS patients were at the moment of testing in the KS phase and not in the acute Wernicke Encephalopathy. Patients had an extensive history of alcohol abuse, were able to read and speak Dutch and informed consent was obtained from the participant and/or legal representative. People with other neurological disorders (e.g., brain tumor or a stroke) or acute psychiatric conditions (e.g., major depression, psychosis, etc.) that could interfere with the interview were exempted from participating in this study. The exclusion criteria were checked by looking at the medical charts. Patients that fulfilled the criteria were asked to participate in this project.

Participants

Thirty patients (11 women and 19 men) diagnosed with KS were included in the current study. The mean age was 65.7 years (SD = 7.3). The education levels that were present in the sample were primary school (n = 1), junior general secondary education (n = 13), senior general secondary school (n = 1), post-secondary vocational education (n = 9), higher vocational education (n = 3) and university (n = 3). All patients were residing in the long-term care facility “Korsakoff Center of Expertise Slingedael,” and were previously admitted to a post-acute rehabilitation facility for Wernicke Encephalopathy. On average, patients received visits of family members or friends once per month, indicative of relative social isolation (Oudman et al., 2018).

Procedure

Two self-report personality questionnaires (DCPI, SDPI) were administered by the same psychologist on the same day in a counterbalanced order. The total procedure took approximately 35–45 minutes. A proxy (e.g., child/(ex) partner/brother/sister) was contacted for each KS patient and asked to participate in the study. The total sample of this study consisted of a total amount of 30 participants. The questionnaire, measuring the premorbid personality traits (HAP) of the KS patient, was orally administered with the informant of the patient. Conditions of inclusion were that the informant knew the patient for a long-term period and across different situations (e.g., work, leisure, and private) and was able to speak Dutch. After this, 30 informants that only had limited acquaintance with the patient (nursing staff) were asked to fill out a questionnaire measuring current personality traits (HAP-t). Administration of the HAP(t) took approximately 10 minutes. Ethical approval of this study was obtained by the ethical committee of social sciences Utrecht University.

Instruments

Dutch Clinical Personality Inventory (DCPI)

The Dutch Clinical Personality Inventory (DCPI; Barelds & Luteijn, 2015) measures personality traits. The questionnaire consists of 120 items (20 items per scale) and is answered on a three-point Likert Scale (False, ?, True). The questionnaire consists of six scales: “Negativism” (NEG; e.g., “I am often dissatisfied”), “Somatization” (SOM; e.g., “I am quickly tired”), “Shyness” (SHY; e.g., “Sometimes I wish that I wasn’t this shy”), “Severe Psychopathology” (PSY; e.g., “Some people hope that something worse will happen to me”) “Extraversion” (EXT; e.g., “Sometimes, I suddenly have a lot of energy”) and Narcissism (NAR; e.g., “I am a very special person”). The reliability and validity of the DCPI are satisfactory and the questionnaire can be used for scientific purposes. Internal consistency reliability is considered good (Barendse et al., 2013). The scores of the KS patients were compared with scores of a representative sample of the general population (Barelds & Luteijn, 2015). The mean scale scores on the questionnaires could be classified as extremely low, low, below average, average, above average, high, or extremely high, in comparison to the norm group. We used the classification system described in the manual.

Theory Driven Profile Interpretation (TDPI) of the DCPI

Another way to interpret the results of the DCPI is by looking at the structural personality organization (PO) assessed by the so-called Theory Driven Profile Interpretation (TDPI). Structural diagnosis is an important addition to the assessment of psychopathology, especially in predicting the course of a disorder and treatment response. The TDPI is based on the idea that the scales on the DCPI should not be interpreted separately but instead should be combined to construct personality organizations, which are formed early in life and are described in detail by L. Eurelings-Bontekoe and Snellen (2021). These organizations can be interpreted by the tripartite model of personality Kernberg (1984). He described the neurotic- (NPO), borderline- (BPO), and psychotic organization (PPO). The PPO is the weakest of the three POs. The main characteristic of the PPO is that reality testing is severely damaged and there are no boundaries between self and others. They are unable to cope with stress, they only use primitive defense mechanisms, have difficulties functioning in society and do not have illness-insight. This organization is followed by the BPO. In this organization, reality testing is intact. However, they suffer from identity diffusion, which means that the sense of self and others is fragmented. Due to this diffusion, they have an inconsistent view of themselves and others. Furthermore, they lack regulating affect mechanisms and make use of primitive defenses. One of the main primitive defenses used by people with a BPO is “splitting.” For these people things are “all good” or “all bad,” they constantly flip back and forth between these two opinions. The NPO is the healthiest organization of the three. Reality testing is intact, identity is integrated and these people are able to cope with stress (E. H. Eurelings-Bontekoe et al., 2008, 2009).

Shortened Dutch Personality Inventory (SDPI)

The SDPI (Barelds et al., 2018) is a shorter version of the most frequently used Dutch personality questionnaire, the NPV-2-R (Barelds et al., 2007). The questionnaire measures personality traits, consists of 70 items (10 items per scale), and can be answered on a 3-point Likert scale (True, ?, False). The questionnaire consists of seven scales: “Neuroticism” (NEU; e.g., “Often, I am feeling sad”), “Social Anxiety” (SA; e.g., “In a group of unknown people, I feel good”), “Rigidity” (RG; e.g., “I like to do my work precisely”), “Hostility” (HS; e.g., “I think most people are trustworthy”), “Complacency” (CO; e.g., “Often, I only think of myself”), “Dominance” (DO; e.g., “I like to make decisions on behalf of others”), and “Self-esteem” (SE; e.g., “I often feel insecure”). The reliability and validity of the SDPI are satisfactory and the questionnaire can be used for scientific purposes. The scores of the KS patients were compared with the scores of a representative sample of the general population (Barelds et al., 2014). The mean scale scores on the questionnaires could be classified as extremely low, low, average, high, or extremely high, in comparison to the norm group. We used the classification system described in the manual.

Hetero-Anamnestic Personality questionnaire (HAP)

In addition to self-reports, earlier studies examining severe neuropsychiatric disorders such as severe depression, psychosis, and dementia patients frequently applied proxy-based personality questionnaires. The reason to do so is that the severity of cognitive problems hampers self-report (van den Broeck et al., 2010). Due to the severe cognitive symptoms of KS patients, this study will also use proxy-based questionnaires (Egger et al., 2002). Within long-term care for dementia patients a frequently applied observant scale is the Hetero-Anamnestic Personality Questionnaire (HAP; Barendse et al., 2013). The HAP can guide professionals working in long-term care by understanding behavioral difficulties in complex patient groups. The HAP (version 2; Barendse et al., 2012) is a short informant-questionnaire which measures personality traits and characteristics of personality disorders. There are two different versions of the questionnaire, the first one assesses premorbid personality traits (HAP) and the second one assesses current personality traits (HAP-t). The HAP can be filled in by a person close to the patient (e.g., child/(ex) partner/brother/sister) and the HAP-t can be filled in by nursing staff. Both questionnaires consist of 62 items with 10 clinically relevant scales: Socially Avoidant Behaviour (SA; e.g., “Had/has a tendency to avoid social situations”), Uncertain Behavior (UNC; e.g., “Behave(s/d) insecure”), Vulnerability in interpersonal relationships (VUL; e.g., “Often feels/felt undervalued or disappointed”), Somatizing Behavior (SOM; e.g., “Often has/had vague physical complaints”), Disorderly Behavior (DIS; e.g., “Is/was accurate”), Rigid Behavior (RIG; e.g., “Has/had difficulties with changing plans”), Perfectionist Behavior (PERF; e.g., “Had/has no rest as long as his work was/is not finished”), Antagonistic Behavior (ANT; e.g., “Was/Is suspicious”), Self-satisfied Behavior (SELF; e.g., “Had/has a tendency to intimidate others”), Unpredictable and Impulsive Behavior (UNP; e.g., “His behaviour was/is unpredictable”). The reliability and validity of the HAP (t) are satisfactory/good on all criteria (Thissen & Barendse, 2019).

Design and data-analysis

Seven HAP questionnaires were missing because proxies could not be tracked (5 times) or because they did not want to participate in the study (2 times). For one participant, the HAP was excluded due to a high score on the “Inconsistent Answers” scale, suggesting that the questionnaire was answered inconsistently and the profile was therefore invalid (Thissen & Barendse, 2019). In total, 22 HAP- and 30 HAP-t questionnaires could be used for the analyzes. First, the scores on the DCPI were interpreted according to the Theory Driven Profile Interpretation (TDPI). The distribution of POs in KS patients were compared to those of a group of 261 psychiatric patients (L. Eurelings-Bontekoe & Snellen, 2021). A Chi-Square Test of Contingencies was performed to examine the differences in the distribution of personality organizations between the KS patients group and the psychiatric patients group. In order to establish the self-reported personality traits of KS patients, descriptives were used to calculate the mean scale scores of the KS patient group on the DCPI and SDPI. The manual was used to compare the mean scores of the KS patients to the mean scores of the general population. Percentage distributions of all scales were calculated to determine the percentage KS patients in each category of the SDPI (extremely low, low, average, high, extremely high) and DCPI (extremely low, low, below average, average, above average, high, extremely high). One sample t-tests were used to examine the difference between the sample mean scores and the mean scores of the norm group. For the scales that violated the normality assumption, no statistical analyses could be performed due to the absence of the median of the norm group. The mean scores on the scales “Neuroticism” and “Dominance” on the SDPI were calculated separately for men and women because the norms were separated for these two groups. To establish the differences between current- and premorbid personality traits in KS patients, paired samples t-tests were used on the similar scale scores of the HAP and HAP-t. Because the normality assumption was violated for the difference scores on the scales ANT, SELF and UNP, non-parametric Wilcoxon Signed Rank tests were used for these scales. All matching scales were paired with each other. Indications of characteristics of personality disorders on the HAP and HAP-t were clustered in cluster A (paranoid-, schizoid- and schizotypical personality disorder), cluster B (anti-social-, borderline-, histrionic- and narcissistic personality disorder) and cluster C (avoidant-, dependent- and obsessive-compulsive personality disorder) and prevalence rates were calculated (Magnavita, 2003). A Chi-Square Test of Contingencies was performed to examine the difference between the KS patients group and a representative sample of the psychiatric population (Lenzenweger et al., 2007). A p-value of < .05 was considered statistically significant. All analyses were performed in IBM-SPSS Statistics 28.

Results

Theory Driven Profile Interpretation (TDPI) of personality organizations

The underlying personality organizations on the DCPI were indexed based on the dynamic theory-driven Kernberg personality inventory. Pie Charts 1 and 2 visualize the distribution of the personality organizations. 13% (n = 4) of the KS patients could be grouped in the neurotic personality organization in comparison to 30% (n = 77) of the psychiatric patients in a normative sample of 261 individuals. 67% (n = 20) of the KS patients could be grouped in the borderline personality organization in comparison to 64% (n = 168) of the psychiatric patients. 20% (n = 6) of the KS patients could be grouped in the psychotic personality organization in comparison to 6% (n = 16) of the psychiatric patients. The chi-square test of contingencies was statistically significant, Χ² (2, N = 291) = 9.40, p = .009, suggesting a difference in grouping between the three main Kernberg personality organizations. The difference between the KS patients group and the psychiatric patients group on the distribution of personality organizations was relatively small, ϕ = .18 (Allen & Bennett, 2014). To summarize, these results show that the distribution of personality organizations between KS patients and psychiatric patients is different. KS patients display less frequently the (most stable) neurotic personality organization and more frequently the (weak) psychotic organization compared to psychiatric patients. These results could suggest more weak and vulnerable personality organizations in KS patients compared to psychiatric patients.

Pie Chart 1. Distribution of personality organizations KS patients (N = 30).

Pie Chart 2. Distribution of personality organizations psychiatric patients (N = 261).

Self-reported personality traits (DCPI)

Table 1. Comparison of DCPI scales between KS patients and general population.

Scales on DCPI (n = 30)	Mean KS patients (SD)	Mean General population	Mean classification	Percentage (%) KS patients Extremely low	Percentage (%) KS patients Low	Percentage (%) KS patients Below average	Percentage (%) KS patients Average	Percentage (%) KS patients Above average	Percentage (%) KS patients High	Percentage (%) KS patients Extremely high
NEG	12.9 (6.6)	9.5	Above average*	-	3.33	23.33	23.33	10.00	40.00	-
SOM	10.8 (7.6)	9.9	Average	-	3.33	33.33	36.67	10.00	10.00	6.67
SHY	16.7 (8.8)	14.5	Above average	-	13.33	16.67	20.00	23.33	20.00	6.67
PSY	4.2 (5.3)	4.1	Average	-	-	30.00	40.00	13.33	10.00	6.67
EXT	16.0 (7.8)	19.1	Average*	10.71	14.29	14.29	32.14	10.71	17.86	-
NAR	11.6 (9.1)	17.2	Below average*	43.44	20.00	6.67	3.33	3.33	20.00	3.33

Notes: *= significant p < .05 (two-sided); SD = Standard Deviation; NEG = Negativism; SOM = Somatization; SHY = Shyness; PSY = Psychopathology; EXT = Extraversion; NAR = Narcisissm.

The DCPI assessed personality traits as reported by the KS patient. Table 1 displays the mean scores of KS patients on the DCPI in comparison to the general population. In comparison to the general population, KS patients scored above average on Negativism and Shyness and below average on Narcissism. KS patients scored average on all other subscales. The above-average score on Negativism in KS patients suggests a tendency toward a negative and dissatisfied attitude toward others. High scores on Negativism also indicate a tendency toward complaining if discomfort is experienced. The above-average score on Shyness suggests a tendency to be insecure, socially awkward, expectant and that they do not feel at ease in the company of strangers. KS patients reported a below-average score on Narcissism, suggesting that they do not think very highly of themselves and are not controlling or bossy.

Self-reported KS personality traits (SDPI)

The SDPI assessed personality traits as reported by the KS patient. Table 2 visualizes the mean scores of KS patients on the SDPI in comparison to the general population. The KS patient group scored high on Complacency, and the KS patient male group scored high on Neuroticism in comparison to the general population. A high score on Complacency suggests that KS patients reported more complacent, self-satisfactory, and egocentric behavior than the general population. A high score on Neuroticism suggests that KS patient males, but not females, self-reported more nervous, vulnerable, and emotionally unstable behavior in comparison to the general population.

Table 2. Comparison of SDPI scales between KS patients and general population.

Scales on SDPI (n = 30) Male (n = 19), Female (n = 11)	Mean KS patients (SD)	Mean General population	Mean classification	Percentage (%) KS patients Extremely Low	Percentage (%) KS patients Low	Percentage (%) KS patients Average	Percentage (%) KS patients High	Percentage (%) KS patients Extremely high
NEU Male NEU Female	6.5 (6.5) 6.3 (5.9)	3.8 6.4	High Average		- 18.18	40.00 54.55	30 9.09	30 18.18
SA	7.8 (5.3)	7.5	Average	10.00	3.33	46.67	26.67	13.33
RG	15 (4.8)	12.3	Average	6.06	3.03	30.30	18.18	42.42
HS	10.7 (5.6)	9.5	Average	8.33	13.89	27.78	19.44	30.56
CO	7.1 (4.1)	4.5	High*	3.45	-	37.93	20.69	51.72
DO – Male DO – Female	9.2 (6.7) 5.2 (4.0)	10.6 8.1	Average Average	52.63 36.36	15.79 36.36	10.53 9.09	- 18.18	21.05 -

Notes: *= significant p < .05 (two-sided); SD = Standard Deviation; SDPI = Shortened Dutch Personality questionnaire; KS = Korsakoff Syndrome; NEU = Neuroticism, SA = Social Anxiety; RG = Rigidity; HS = Hostility; CO = Complacency; DO = Dominance; SE = Self-Esteem.

Difference between current- and premorbid informant-reported personality traits

The HAP-t was used to assess the current informant-reported personality traits and the HAP to assess the premorbid informant-reported personality traits of KS patients. Table 3 displays the mean percentages and the differences between informant-reported current- and premorbid personality traits. All mean percentages on the scales of HAP-t and HAP were around the average (50%) which means that some traits of that scale could be attributed to the KS patients. Current personality traits (HAP-t) that were most frequently observed by the nursing staff were as follows: rigid behavior, antagonistic behavior and socially avoidant behavior. Premorbid personality traits (HAP) that were most frequently observed by proxies were as follows: vulnerability in interpersonal relationships, antagonistic- and perfectionist behavior.

Table 3. Mean percentages and differences between current- and premorbid personality traits KS patients.

	HAPt (n = 22) Mean percentage of maximum (100%) score (SD)	HAP (n = 22) Mean percentage of maximum (100%) score (SD)	Test statistic	p-value	Effect size (d)
SOC	44.4 (27.0)	27.3 (25.1)	t = −2.083	.050	0.66
UNC	38.3 (27.5)	35.9 (37.6)	t = −.282	.780	0.07
VUL	42.8 (25.2)	51.9 (29.8)	t = 1.657	.112	−0.33
SOM	23.1 (28.2)	20.6 (29.8)	t = −.311	.759	0.09
DIS	42.3 (34.2)	42.2 (31.7)	t = −.011	.991	0.00
RIG	55.6 (29.9)	38.9 (23.5)	t = 2.495	.021*	0.63
PERF	34.1 (33.7)	44.6 (28.6)	t = 1.207	.241	−0.34
ANT	46.7 (29.6)	49.1 (25.1)	Z = −.121	.896	−0.09
SELF	42.1 (37.0)	32.7 (38.6)	Z = 1.457	.145	0.25
UNP	40.7 (30.4)	44.0 (31.1)	Z = −.037	.970	−0.11

Notes: *= significant p < .05 (two-sided); SD = Standard Deviation; d = Cohen’s d (effect size); KS = Korsakoff Syndrome; HAP(t) = Hetero Anamnestic Personality questionnaire (current traits); SOC = Socially Avoidant Behaviour; UNC = Uncertain Behaviour; VUL = Vulnerability in interpersonal relationships; SOM = Somatizing Behaviour; DIS = Disorderly Behaviour; RIG = Rigid Behaviour; PERF = Perfectionist Behaviour; ANT = Antagonistic Behaviour; SELF = Self-satisfied Behaviour; UNP = Unpredictable and Impulsive Behaviour.

One paired samples t-test demonstrated that current rigid behavior was significantly higher compared to the premorbid rigid behavior, t(21) = 2.50, p = .021. Cohen’s d for this test was 0.63, which can be described as a medium effect. Moreover, the analysis showed a trend toward a significantly higher score on current socially avoidant behavior compared to premorbid socially avoidant behavior (p = .05), suggesting worse social functioning in the current state compared to the premorbid state. All other scores did not differ significantly between current- and premorbid personality traits. To summarize, informants observed that rigid behavior and socially avoidant behavior increased as a result of KS. Both central traits were already present prior to the KS diagnosis.

Prevalence rates of characteristics of personality disorders in KS patients, based on the observational scales (HAP and HAP-t), compared to estimated prevalence rates of a representative sample of the American population

The prevalence rates of personality disorder characteristics in KS patients and the estimated prevalence rates in a non-clinical sample were examined to investigate differences between these groups. Table 4 visualizes the prevalence rates of informant-reported personality disorders characteristics in KS patients on the HAP (N = 22) and HAP-t (N = 30). These prevalence rates were compared to prevalence estimates based on data from the National Comorbidity Survey Replication (NCS-R; Lenzenweger et al., 2007), which consists of a representative group of the American population (N = 5692). The chi-square test of contingencies was statistically significant, Χ² (4, N = 776) = 33.29, p < .001, although the difference between the KS patients group and general population on the clusters of personality disorders was quite small, φ_c = .21. The NCS-R group demonstrated a higher estimated prevalence rate on cluster A personality disorders compared to KS patients on the HAP and HAP-t. However, KS patients demonstrated higher prevalence rates on clusters B and C compared to the NCS-R group. In sum, based on these observations it seems that KS patients more frequently display signs of personality disorders compared to people in the general population, especially in clusters B and C.

Table 4. Prevalence of clusters personality disorders in KS patients and prevalence estimates in a non-clinical sample.

	% KS patients (HAP)	% KS patients (HAP-t)	% NCS-R
Cluster A	4.6	3.3	5.7
Cluster B	31.8	16.7	1.5
Cluster C	27.3	16.7	6.0

Notes: KS = Korsakoff Syndrome; HAP(t) = = Hetero Anamnestic Personality questionnaire (current traits); NCS-R = National Comorbidity Survey Replication.

Discussion

It is commonly known that KS patients suffer from severe neuropsychiatric symptomatology and as a consequence display behavioral and interpersonal problems. Previous research on personality and mental diseases pointed out that knowledge of personality is important to determine the direction of treatment and develop new treatment strategies for complex psychiatric populations. Despite this importance, the personality traits of KS patients have not been systematically studied yet and the currently available literature is inconclusive. Therefore, we examined the personality of KS patients by using both self-report- and informant-report questionnaires.

Premorbid personality organizations and traits in KS patients

In the current study, we examined whether KS patients have certain premorbid personality organizations that put them at risk to develop severe mental illnesses later in life. Based on the dynamic theory-driven profile interpretation (TDPI) of personality organizations, which measures underlying personality organizations based on the results of the DCPI, we found that the distribution of Kernberg personality organizations of KS patients are different from those reported for psychiatric patients. The neurotic personality organization is less frequently present and the psychotic personality organization is more frequently present in KS patients, suggesting less developed personality organizations in KS patients compared to the psychiatric population. This finding is not remarkable because concurrent with people with a psychotic personality organization, KS patients are vulnerable regarding societal, social, and occupational demands, often do not respond to treatment, and globally lack insight into their problems. Moreover, similar to other people with a psychotic personality organization, KS patients tend to live in social isolation and make use of splitting as a primitive defense mechanism (E. H. Eurelings-Bontekoe et al., 2008; Gerridzen et al., 2018; McWilliams, 2020; Oudman et al., 2018; van den Hooff, 2022). Individuals with a weak and vulnerable personality organization are at risk for developing severe mental illnesses which could indicate that patients already had a certain premorbid vulnerability to develop severe alcohol abuse leading to KS (E. H. Eurelings-Bontekoe et al., 2008). Our finding that problems in early life increases the risk for developing severe mental illness later in life is in line with Vlot (2023), which found that premorbid delinquency is associated with the development of KS later in life. Many patients that later receive the diagnosis of KS suffer from childhood abuse or other adverse childhood experiences that put them at risk for the development of personality problems. It would therefore be relevant to retrospectively investigate the existence of traumatic events in patients diagnosed with KS, also taking into account diminished capacities to retrieve autobiographical memories (El Haj & Nandrino, 2017).

To get a more complete picture of the premorbid personality of KS patients, we used an informant questionnaire (the HAP), which was filled in by a proxy, to assess premorbid personality disorders characteristics and traits. In comparison to normative data of the general population, proxy-based reports of behaviors linked to personality disorders indicate that KS patients show more premorbid traits of cluster B- and C-personality disorders. Specifically, KS patients tended to show overly emotional-, unpredictable-, dramatic-, anxious- and fearful behavior. These observed personality problems by informants are in line with our finding that KS patients have an underlying vulnerable and weak personality organization leading to adverse personality development later in life, and with the finding that an increase in personality disorder characteristics is concurrent with addiction behavior (E. H. Eurelings-Bontekoe et al., 2008; Fehrman et al., 2017; Fromme et al., 2004). Interestingly, the proxy responses did not show many specific personality problems that are frequently present in KS, but suggested a complex pattern of personality traits with variable distribution.

Current personality traits in KS patients

We used two self-report questionnaires, the DCPI and SDPI, to evaluate the current personality traits in a representative sample of KS patients. In line with Egger et al. (2002), we did find an average score on most scales measuring self-reported personality traits, suggesting that the patients report themselves as healthy individuals. This is opposed to what we would have expected based on the severe neuropsychiatric symptoms they suffer from, and in contrast to the personality problems observed by informants. In contrast to Egger et al. (2002), our results indicate that KS patients could be distinguished by increased self-reported negativism, shyness, and complacency compared to the general population. Additionally, KS males, but not females, could be distinguished by increased levels of self-reported neuroticism compared to the general population. Based on earlier findings in alcohol use disorder patients, these findings are in line with our expectation that KS patients would score higher on these scales compared to the general population (Gerridzen et al., 2017; van Dam et al., 2020; Wilson et al., 2012). We also found a decrease in self-reported narcissism in KS patients compared to the general population. This was against our expectations because narcissistic tendencies are often observed in clinical practice and the literature mentions that KS patients often display aggression, apathy, and arrogant behavior (Gerridzen et al., 2017). An explanation for this result could be that the questions on this scale were answered in a socially desirable manner because they were quite extreme (e.g., “I am a very special person,” and “I like to watch myself in the mirror”). Moreover, clinical narcissism is associated with the denial to have self-reported narcissistic traits (Cooper et al., 2012). Also, the aforementioned weak personality organization in KS patients could reflect a lack of insight into their own personality, which is often found in individuals with weaker personality organizations.

Proxy- and nursing staff reports on current personality traits in KS

Further, we examined the current personality by using an informant questionnaire (HAP-t), which was filled in by the nursing staff. Similar to what proxies reported, nursing staff reported that KS patients tended to show overly emotional-, unpredictable-, dramatic-, anxious- and fearful behavior. Nursing staff also did not report many personality problems based on the separate scale scores on the HAP-t. Of importance, proxies observed some premorbid rigid behavior and socially avoidant behavior, but informants observed even more increased levels on these scales as a consequence of KS. These results suggest that the detrimental effects of KS are seen in rigid behavior and socially avoidant behavior. Both symptoms have been reported earlier in the KS literature on the severity of (social) cognitive impairments (Drost et al., 2018) and loneliness (Oudman et al., 2018), but also in the literature on executive disorders in KS reflecting switching problems (Moerman van Den Brink et al., 2019). Our finding that these symptoms are not only observed in cognitive tests but also reported in care staff reports on personality adds to the literature about KS. These findings also show that the HAP(t) is a valuable addition to self-report questionnaires when examining personality traits in KS patients because KS patients did not report rigid- and socially avoidant behavior themselves.

Relationship between current- and premorbid personality traits

We also analyzed the relationship between informant-reported current- and premorbid personality traits. The results indicated that when nursing staff observed an increase on several current personality traits, proxies indicated that these increases were also observed premorbid to the disease, and vice versa. This finding suggests that these personality traits in KS patients are relatively stable over time. This is also in line with research on alcoholism, showing personality traits such as neuroticism and impulsivity before and during the addiction (Le Bon et al., 2004). These findings underline the complexity of the syndrome because they suggest that KS patients lack self-insight and tend to blame others but not themselves.

The status of Theory-Driven Profile Interpretation (TDPI)

In this research, we embraced the theory-driven profile interpretation (TDPI) proposed by L. Eurelings-Bontekoe and Snellen (2021). Although many clinicians and researchers in facilities for patients diagnosed with personality disorders adopt TDPI, it has also faced criticism on statistical grounds, such as its reliability and validity (Smid & Kamphuis, 2005). Critics argue that TDPI should not find application in clinical practice due to the lack of empirical evidence and its inability to be deemed psychologically “healthy” (Smid & Kamphuis, 2005). While the debate extends beyond the scope of this manuscript, it is essential to note that TDPI has been subject to criticism as a model for personality interpretation.

Limitations

It is plausible that a number of limitations could have influenced the results that were obtained in this study. A first limitation of this study is the relatively small sample size, possibly leading to relative statistical power issues. We would like to point out that the present study on personality in KS has comparable sample sizes to earlier studies on personality traits in KS. Egger et al. (2002) included 20 KS patients, and S. Walvoort et al. (2016) included 34 KS patients. While it would be of importance to include larger samples of KS patients, the relative rarity of the disorder complicates this factor. A second limitation of this study is that there is no background information about the proxy available. For instance, a spouse might have more long-term information about the past of a patient compared to a child which might have influenced the results. In future studies it is important to gather this information and take differences in proximity to the patient into account.

A limitation worth mentioning is that several scales on the DCPI and SDPI were not normally distributed. Due to this, one sample t-tests could not be performed. We were also not able to perform non-parametric tests because the median of the scales was absent in the manual. A limitation we came across when interpreting the HAP(t) questionnaire was that the cutoff scores of the questionnaires were not available in de manual. Therefore, we could not determine when a score was above- or under the average score in comparison to the norm groups.

Recommendations and clinical implications

Due to the lack of self-insight we recommend including the perspectives of the KS patients themselves and their proxies/nursing staff in future research, in which at least underlying personality organizations and informant-reported personality traits should be measured. In future research, we think it is also relevant to examine the relationship between social cognition and informant-reported socially avoidant- and rigid behavior to examine if social cognition can predict these behaviors. To gather more information about the neuropsychological profile of KS patients, we think it is also relevant to include aspects such as apathy, depression and disinhibition in future studies and correlate this factors with outcomes on personality tests. Our findings suggest an overrepresentation of at-risk personality organizations in KS patients, which are observed in personality problems by the environment later in life. Therefore, we recommend to investigate the relationship between underdeveloped personality organizations and the development of severe neuropsychiatric disorders later in life.

In clinical practice, we think it is important to develop training programs for the medical team where interventions are explained on how to deal with rigid- and socially avoidant behavior. This might be helpful because these problems require an intensively structured environment, which includes for instance fixed daily routines and establishing clear agreements. Further, specific attention for increasing guided social interaction between patients and with the nursing staff might be helpful. Next to this, because patients with lower personality organizations can be highly provocative, it is important to be aware of the emotional effect that these patients can have on others and by explaining concepts such as “splitting.” Supportive interventions such as limit setting, clarifications of tasks and roles, undermining maladaptive primitive defenses and reinforcing adaptive defenses and focusing on a strong therapeutic alliance may be helpful in dealing with these patients (Gordon et al., 2019).

Conclusion

In conclusion, this study has provided new insights into the personality of KS patients, although caution is warranted given the study’s limited sample size and the novelty of the research field. While KS patients generally reported few personality issues and exhibited overall traits consistent with those of healthy individuals, our analysis revealed a tendency toward a weak and vulnerable personality organization among this group. Informants reported an increase in current socially avoidant- and rigid behavior and observed current and premorbid personality problems such as emotional-, unpredictable-, dramatic-, anxious and fearful behavior. We found differences between self-reported personality traits and informant-reported personality traits, which indicates a lack of self-insight possibly caused by weak personality organizations in KS patients. Changes in the personality caused by KS, such as rigid behavior and socially avoidant behavior, require an structured environment, with fixed daily routines and clear agreements, and specific attention for increasing guided social interaction between patients and with the nursing staff. However, further research with larger sample sizes is necessary to confirm and expand upon these initial findings. We also recommend training programs for the medical team focussed on psycho-education and supportive interventions for patients with complex personality problems.

Acknowledgments

The content of this manuscript has not been published elsewhere. All authors contributed significantly and are in agreement with the content of the manuscript.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data is available upon request. We did not preregister the research plan in an institutional registry.

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.