INTRODUCTION
Robert Cloninger’s concept of temperament and its relationship to alcohol dependence has practical applications, as has been confirmed by numerous studies over many years [1, 2]. The available questionnaire examining temperament and character traits, e.g., the Temperament and Character Inventory (TCI), can easily be applied in clinical practice. This questionnaire assesses the expression of temperament traits (because they reflect the functioning of genetically determined neurotransmitter systems according to Cloninger’s conception) like Novelty Seeking, Harm Avoidance, Reward Dependence and Persistence [1].
Although substantial research has already been carried out on this topic, it is essential to explore further the determinants of different appetitive behaviours, including factors that protect or predispose an individual to addiction [3, 4].
Appetitive behaviours, in addition to the seeking of psychoactive substances, can pose additional health risks to the person with an addiction [5, 6]. In addition to seeking psychoactive substances (where the motivator is, for example, a craving for alcohol), appetitive behaviour can include food intake (where the motivator is hunger) or taking medication to obtain alternative sensations to alcohol (e.g., feeling better following pain relief) [7]. A consequence of impulsive food intake can be a nutritional disorder (e.g., obesity) [8], while over-frequent painkiller intake bears with it the risk of side effects [9].
Individual temperament traits have been linked to neurotransmission-determinant genes and their expression. Studies have demonstrated that the Val158Met polymorphism of the COMT gene is a marker of depressive response to stress and impulsiveness, shown to correlate with the Novelty Seeking temperament trait [10]. The Taq1A polymorphism of the DRD2 gene is a marker of increased alcohol consumption and impulsivity, which correlates with Novelty Seeking, Harm Avoidance or persistent temperament traits [11]. In contrast, the SLC6A4 gene 5-HTTLPR polymorphism, which encodes a serotonin transporter, may be associated with harm avoidance and temperament traits of vulnerability to depression. However, the clinical variable association depends on the gene polymorphism variant (e.g., 5-HTTLPR short/short) [12].
It is possible that, in the context of appetitive behaviour, apart from genetic factors determining temperamental traits (e.g. Novelty Seeking, Reward Dependence, Harm Avoidance), biochemical factors may play an important role, e.g., orexin and neuropeptide Y as potent orexigenic hormones [13]. Orexin is responsible for an increase in the food-seeking need [14]. Neuropeptide Y is also responsible for regulating food intake [13].
The study aimed to assess the association of the temperamental traits and biochemical variables that determine appetitive behaviour with alcohol-dependent individuals’ health status, food preferences and anthropometric parameters.
METHODS
Survey organisation and sample characteristics
The total sample consisted of 154 individuals (140 men and 14 women – it was decided to leave the women’s group out of the study following the verification of the temperamental traits analysed regarding gender). Recruitment was conducted among alcohol-dependent patients from several alcohol rehab centres in the Kujawsko-Pomorskie, Łódzkie and Podlaskie regions of Poland. The study was conducted on patients with at least 2 weeks of abstinence in the first 2 weeks of treatment (it was the first point of study, symbol “I”). Some variables, like the severity of symptoms of depressive and anthropometric parameters, were also assessed 4 weeks after the first examination (i.e., the 6th week of hospitalisation, the second point of study, symbol “II”).
The conditions for participation in the study included meeting at least one inclusion criterion (18 years of age, having no other addictions except alcohol and nicotine dependence); the exclusion criteria were noticeable cognitive deficits, chronic metabolic diseases like diabetes or active viral or bacterial infections, including liver dysfunction of an infectious nature, and being unable to provide informed consent.
Applied methods
The research was conducted using face-to-face interviews. Specialist researchers carried out the questionnaire-based interviews and measurements. Laboratory tests were performed in medical facilities. The interview focused on basic socio-demographic data such as age and gender and diet (number of days per month on which a particular type of food was consumed, such as meat in the category of red meat and poultry, dairy products, sweets and fruit). The assessment of dietary preferences referred directly to the variables studied, such as the biochemical and temperamental determinants of appetitive behaviour and, consequently, nutritional status (anthropometric parameters).
In addition to the interview, the patient’s length of hospitalisation and their medical records were analysed for their intake of painkillers like paracetamol, ketoprofen, metamizole and ibuprofen (the data here pertained to 50 patients).
Questionnaires used
The Temperament and Character Inventory (TCI) questionnaire is a set of scales for the measurement of temperament and character dimensions according to Robert Cloninger’s psychobiological concept [15]. The questionnaire in the Polish adaptation of E. Hornowska [16] was used, consisting of 240 questions which, after the application of a key, define temperamental areas (Novelty Seeking and its sub-dimensions – cognitive curiosity, impulsivity, extraversion, disorderedness; Harm Avoidance, and its sub-dimensions – pessimism, fear of uncertainty, social anxiety; Reward Dependence and its sub-dimensions – sentimentality, attachment, dependence) and Perseverance. The above dimensions formed the basis of the methods of the present study.
The questionnaire also assesses character traits like self-directednes and its with sub-dimensions – responsibility, purposeful behaviour, resourcefulness, self-acceptance, good habits; willingness to cooperate, including social acceptance/tolerance, empathy, willingness to help, forbearance, integrated conscience sub-dimensions; ability to detach from self/transcendent self, and its sub-dimensions the creative transcending of self, transpersonal identification, and acceptance of spirituality. However, these dimensions were not included in this study.
Each dimension and sub-dimension is summarised through a spreadsheet algorithm. The results of this study’s individual dimensions and sub-dimensions were compared with those of the control group and interpreted by a clinical psychologist. The higher the score of a given subscale, the more expressed the measured trait is in the subject [15, 16].
The Short Alcohol Dependence Data Questionnaire (SADD) is used to assess the depth of alcohol dependence. The scale consists of 15 questions. Four response options are possible. The successive response options correspond to degrees of depth of addiction and are therefore scored in ascending order from 0 to 3 points. The depth of addiction is assessed based on the total score obtained. The following criteria are adopted: 1-9 points – mild depth of addiction; 10-19 points – moderate; 20-45 points – deep. The scale has achieved good psychometric properties in studies [17, 18].
Assessment of alcohol craving and pre-meal hunger
The Penn Alcohol Craving Scale (PACS) consists of five test items. Three questions relate to the frequency, intensity and duration of craving; one measures the ability to resist temptation when drinking is possible, and another estimates the degree of overall alcohol craving over the past week. Responses are given on a scale of 0-6. The scale has good psychometric properties and is very often used in studies of alcohol cravings. Furthermore, it can predict, better than other methods, the risk of relapse during treatment. The score range 0-3 indicates a low intensity of craving for alcohol; 4-9 is average; and 10 and above is a high intensity of craving for alcohol – Polish adaptation [19, 20].
The Yale-Brown Obsessive Compulsive Drinking Scale (Yale-Brown Obsessive Compulsive Scale Modified to Reflect Obsessions and Compulsions Related to Heavy Drinking, Y-BOCS). It is used to measure thoughts and behaviours regarding obsessions and compulsions about drinking alcohol that occur in abusers and addicts. The scale consists of 10 statements, of which five relate to obsessive thoughts about alcohol (e.g., their intensity, frequency) and a further five relate to compulsive drinking behaviour (e.g., subjective feeling of loss of control over drinking, effort put into abstaining from drinking). According to the scale’s authors, the total responses can be summed up, and the results can be analysed in two subscales [21, 22].
The hunger scale (HS), a questionnaire constructed by the authors, consists of 15 questions describing feelings of hunger and eating behaviour, symptoms co-occurring with feelings of hunger, food intake control and feelings of satiety. For example: Q1 – “When I wake up, I constantly think about eating something”, Q2 – “Immediately after waking up I have to eat something”, Q3 – “I don’t wait until the morning for breakfast, I eat something at night because of being very hungry”, Q4 – “When I don’t eat for a long time I grab anything to get full”, Q5 – “When I am hungry, I feel weak and think I am going to faint”, Q6 – “When I am hungry my hands tremble”, Q7 – “When I am hungry I am irritable”, Q8 – “When a strong uncontrollable urge to eat something suddenly comes over me, I have to eat something immediately”, Q9 – “When I start to eat, I eat quickly and in large quantities”, Q10 – “The need to eat something is so strong and frequent that I snack between main meals”, Q11 – “If I am hungry it disturbs me so much that I cannot concentrate on anything other than my food”, Q12 – “Often I am so hungry and if I cannot eat something, I light up a cigarette”, Q13 – “Even when I eat something, I have to eat it every now and then because I do not feel full”, Q14 – “Sometimes I am so hungry I eat something despite internal resolutions not to or the rules”, Q15 – “Even though I am not that hungry and I see something I like to eat, the hunger builds up so much that I cannot stop myself from buying/eating it”.
The patient assesses the last week prior to the study by answering the questions on a scale of 0 to 3, i.e., 0 never happened, 1 only sometimes, 2 often, 3 almost always. The scores of the individual answers are added up. The maximum number of points is 45. The higher the score, the greater the intensity of hunger before meals. The questionnaire has good parametric properties. The reliability of Cronbach’s α coefficient was 0.86.
Biochemical tests
Serum for determining orexin and neuropeptide Y was obtained from venous blood collected from fasting subjects in a dry tube. The blood was centrifuged for 15 minutes at 3500 rpm at 4°C, and the serum was frozen at 80ºC. The concentrations of orexin and neuropeptide Y were assessed by an immunoenzymatic test – ELISA. Reagents from Cloud-Clone Corp., USCN Life Science Inc., were used to determine the concentrations of both parameters (CEA607Hu ELISA for Orexin A (OXA), CEA879Hu ELISA for Neuropeptide Y (NPY)). The study did not refer to the reference values of the above parameters and only compared OXA and NPY concentration values between the patient groups studied.
Assessment of nutritional status using anthropometry
A TANITA MC-780 S MA multi-frequency segmented body composition analyser was used to assess anthropometric variables. The measurement was non-invasive, using the electrical bioimpedance (BIA) method. The device is medically certified and can be used for scientific research. Using the above-mentioned device, the following was determined in the subjects: body fat content, i.e. the percentage of body fat (%Fat Mass, %FM, %), relative body mass index (BMI, kg/m2) based on body weight (kg) and, in addition, body height (cm), lean body mass (Free Fat Mass or FFM is a parameter that refers, among other things, to bone and muscle mass; normal muscle mass can be an indicator of normal protein-calorie nutrition) [23].
Bioethics
This study was conducted according to the guidelines in the Declaration of Helsinki, and all procedures involving human patients were approved by the Bioethics Committee (consent number KB 692/2017). Written informed consent was obtained from all patients.
Statistical analysis
The conformity of the studied variable distribution with the normal distribution was checked.
Tests from the IMAGO6 statistical package (IBM SPSS 26) were used in the study. The level of statistical significance was taken as at least p ≤ 0.05. The Kolomogorov-Smirnov test was used to assess the distribution, which showed that the analysed variables deviated from a normal distribution (part of the variables deviated from a normal distribution, and next, we decided to use non-parametric tests). This determined that Spearman’s rho and Mann-Whitney’s U non-parametric tests were used in the statistical analysis to assess the correlation and compare dichotomic variables. Pearson’s Chi-square test was used to compare the frequencies of variables. It was decided to use the median(s) and ranges of values (minimum – maximum; min. – max.) of the data obtained in the statistical descriptions.
RESULTS
First, the associations of temperament traits measured by the TCI with the results of the PACS, Y-BOCS, SADD and HS questionnaires were assessed (Table 1).
Table 1 data showed that the Novelty-Seeking subscale scores correlate positively with the Y-BOCS, SADD, and HS. The harm avoidance subscale scores correlate with PACS and Y-BOCS. There were low negative correlations between the Reward Dependence subscale and the Y-BOCS and the Persistence subscale with the PACS and Y-BOCS.
Due to the temperament subscales’ weak correlations with the analysed variables, the Novelty Seeking and Harm Avoidance subscales were left for further statistical evaluation.
Further analyses were based on comparing biochemical, clinical and anthropometric variables between patient subgroups according to the strength of the temperament traits in the above-mentioned subscales. A dichotomous division of a given trait was made according to medians for patient subgroups distinguished according to greater or lesser trait expression (lower and higher scores of a given temperament trait) (Tables 2 and 3).
Table 2 shows that the patient group with more expressed Novelty Seeking traits was statistically significantly younger (41 vs. 44 years), with fewer women (3 vs. 11) and with higher Y-BOCS (19.00 vs. 15.00 score), and consumed protein products (meat) more days per month than the group with less expressed traits. The difference in the scores of the other variables was not statistically significant.
Table 1
Correlation of scores of the four of the TCI temperament subscales with PACS (alcohol craving), Y-BOCS (alcohol craving), SADD (deep alcohol dependence) and hunger scale (HS; rate of hunger) scores
Table 2
Comparison of demographic, biochemical, clinical (medication use) and anthropometric variables in alcohol-dependent patients with more or less expressed temperamental traits of the Novelty Seeking TCI subscale
Table 3 shows that the group of patients with more expressed Harm Avoidance traits were statistically significantly older (43 vs. 41 years), with more females (11 vs. 3), took more doses of analgesics during their ward stay (4 vs. 1), had higher serum neuropeptide Y levels determined at week 6 of hospitalisation (1323 vs. 1064 pg/ml) and had higher Y-BOCS scale scores (17.00 vs. 15.00 score) than the group with less expressed Harm Avoidance traits. Notably, the %FM variable was higher in patients with more than less expressed Harm avoidance (20.20 vs. 18.00%), but the difference was close to statistical significance – p = 0.052. Patients with more expressed Harm Avoidance had statistically significantly lower FFM values than patients with less expressed temperament (59.55 vs. 65.40 kg) and consumed sweets less frequently (6 vs. 15 days per month). The difference in the results of the other variables was not statistically significant.
An internal correlation matrix was then prepared and analysed, between the item scores and the overall HS score in the study’s first and second stages. From the matrices obtained, the mean scores were most strongly correlated with the overall scale score, i.e., R at between 0.601 and 0.781, were selected. Subsequently, the mean scores of the selected HS questions were compared in patient subgroups separated according to the median scores of the subscales Novelty Seeking (TCI) and Harm Avoidance (TCI) from both the first and second stages of the study. Only statistically significant results are included in Table 4.
Table 4 shows only the statistically significant differences comparing individual HS items. The results showed that pre-meal hunger was statistically significantly higher in patients with the more expressed temperament trait Novelty Seeking (TCI) at the start of drug treatment, as also shown by the higher scores for questions Q14 and Q15, which indicate a lack of eating control associated with intense pre-meal hunger in these individuals. In addition, individuals with the more expressed Novelty Seeking (TCI) temperamental trait also had severe irritability when experiencing intense pre-meal hunger (refers to the study after one month of abstinence maintenance).
Table 3
Comparison of demographic, biochemical, clinical (medication intake) and anthropometric variables in alcohol-dependent patients with more or less expressed temperamental characteristics of the ‘Harm avoidance’ TCI subscale
DISCUSSION
Subjects who score low on the Novelty Seeking (TCI) subscale are thought to lose their temper less readily, are not so inquisitive, are unenthusiastic, reticent and easily tolerate the monotony of life. They are often seen as orderly, systematic and frugal. In contrast, subjects who score high on this subscale are usually out of control, seeking new experiences, and curious and enthusiastic. They seem full of life and energy. However, they quickly succumb to boredom and exhibit impulsive and disordered behaviour. They have little persistence in action and engage easily with the new and unfamiliar. They tend to have uncontrollable moods and withdraw from situations where needs are unfulfilled, which consequently disrupts their interpersonal relationships [16].
Individuals with low scores on the Harm Avoidance subscale (TCI) appear to be carefree, self-controlled and optimistic, relaxed, outgoing and brave and have dominant self-confidence in social situations. They are perceived as energetic and vigorous. A characteristic of these individuals is confidence in the face of danger. These individuals are prone to feeling overly optimistic and underestimating danger. In contrast, high achievers are cautious, fearful, careful and tense, anxious, nervous, shy, full of doubt, and tend to become discouraged. They may be passive in action and have a pessimistic attitude. They are characterised by being inhibited, shy in social situations, and chronically tired. These individuals are sensitive to criticism and require considerable support and motivational stimuli. They often have a sense of insecurity and anxiety about the future, which causes them to try to plan for situations potentially involving some danger [16].
Table 4
Comparison of the selected questions of hunger scale (HS) in alcohol-dependent patients with more or less expressed temperamental traits of the Novelty Seeking TCI subscale
The first statistical analyses showed a positive correlation of the Novelty Seeking trait (TCI) with the severity of craving for alcohol expressed by the obsessive-compulsive construct (Y-BOCS), with the severity of alcohol dependence (SADD) and, interestingly, with the seriousness of experiencing pre-meal craving (HS). This may indicate that when this temperament trait is more strongly expressed, appetitive needs increase, e.g., the intake of food or alcohol is a source of reinforcement. In contrast, the trait of Harm Avoidance (TCI) positively correlated with the severity of alcohol craving as expressed in the PACS and Y-PACS scale scores. Still, it did not correlate with the severity of pre-meal cravings (HS).
Temperamental risk factors for psychoactive substance use were sought more than 20 years ago, explicitly indicating the Novelty Seeking and Harm Avoidance traits analysed in the present study [24, 25]. It was shown that individuals with high Harm Avoidance and low Novelty Seeking stimulation demand were characterised by anxiety and lowered mood [26]. Following Hornowska [16], it can be added that patients indicated anxiety reduction as the primary ‘motivator’ for taking a drug. According to a 2001 study, individuals with high Harm Seeking scores were more likely to drop out of drug treatment [24].
Low values of the trait Novelty Seeking and high values of Harm Avoidance are characteristic of women [27]. This was confirmed by the present study, which indicated fewer women in the subgroup with higher scores on the Novelty Seeking subscale than those with lower values. Conversely, there were more women in the subgroup with higher values of the Harm Avoidance subscale than in the subgroup of those with lower values of this subscale.
Our further analyses showed that patients with more expressed Harm Avoidance temperamental traits took more painkillers than those with less expressed traits. These individuals may be more sensitive to pain and seek to reduce it to a greater extent or avoid negative sensations altogether. Knaster’s research [28] indicated that pain enhances sensations of anxiety and fatigue. This would be reasonable for the subgroup of patients with a more expressed Harm Avoidance trait, of which anxiety and fatigue are characteristics [16]. Knaster suggests that the Harm Avoidance trait may represent a vulnerability to the occurrence of chronic pain and pain-related anxiety. In addition, the authors highlight the clinical implications in this group of patients, towards whom a targeted therapeutic approach and more intensive therapeutic methods could be more supportive in reducing the negative behaviors connected with Harm Avoidance [28, 29, 30]. The above suggestion is all the more valid concerning alcohol-dependent patients with a more pronounced Harm Avoidance trait in inpatient drug treatment. Moreover, subsequent statistical analyses showed it was this group of patients who experienced increased craving for alcohol (PACS, Y-BOCS), were more profoundly dependent on alcohol (SADD score) and had increased symptoms of depression both at the beginning and in the middle of drug treatment (BDI score).
An interesting phenomenon observed in patients with a more expressed Harm Avoidance temperamental trait was a higher concentration of neuropeptide Y as an orexogenic hormone promoting a sense of hunger. It should be added that this hormone’s expression is increased when the intake of calories is restricted [31]. At the same time, the same subgroup of patients was characterised by a lower value of the FFM anthropometric parameter, indicating an increased energy requirement of this patients’ subgroup (here in the 6th week of hospitalisation) [32] and a lower intake of sweets (high-calorie products). A possible interpretation of the coincidence of the above-mentioned phenomena, including the more frequent intake of analgesics by those with a more expressed Harm Avoidance trait, could be the increased susceptibility of this group of individuals to dental problems. It is possible that toothache, or other oral pain or pain per se, led to reduced calorie intake, including sweets, lowering FFM. Thus, a ‘repair’ mechanism in the form of increased NPY expression was triggered.
After analysing the internal correlation matrix of the HS and its items and their relationship to selected temperament traits, we showed that the statistical differences of the most expressed hunger (hunger) areas of the HS relate to the temperament trait of Novelty Seeking and not Harm Avoidance. We have shown that alcohol-dependent individuals who are characterised by high scores on the Novelty Seeking TCI subscale cannot refrain from eating something (this refers to the beginning of their hospitalisation; results of the comparison of scores of questions Q14 and Q15 of the HS) or are more irritable when feeling hunger before a meal (refer to the second measurement, i.e. after one month of abstinence; comparison of scores of question Q4 of the HS) (Table 4). Past research has shown that the Novelty Seeking trait was correlated with disinhibition (positively correlated) [33]. At the same time, Novelty Seeking (looking for excitement) is associated with type II addiction, according to Cloninger. In subjects with type II alcohol dependence, changes in dopamine neuroconduction (unchanged dopamine transmission) are less pronounced, while there is less serotonin transmission [25]. Lower serotonin transmission is associated with poorer moods, such as lowered mood and a tendency to irritability [34]. The link between irritability and serotonin downregulation is highlighted by research on the efficacy of buspirone as a 5-HT1A receptor agonist in treating irritability in autism [34]. Lower serotonin transmission also leads to greater susceptibility to impulsive behaviour and lack of behavioural control, as confirmed by previous studies by other authors [35-38]. Interesting studies were presented by Tabiś and Poprawa [39], who showed that patients with high frustration of needs (e.g. autonomy and competence) assessed the effectiveness of maintaining abstinence as lower and had a stronger craving for alcohol than the patients with lower feeling of frustration of these needs. Additionally, it was confirmed that loss of control over drinking increases the risk of relapse. It seems that the aforementioned loss of control over drinking and low sense of self-efficacy correlate positively with impulsivity and the need to experience novelty, which is challenging to satisfy, thus triggering a sense of frustration. In one of the most important studies related to the period of the COVID-19 pandemic on the health problems of addicts, it was confirmed that addiction increases the risk of anxiety and depression [40]. The results presented here correspond to the data of Oniszczenko [41], who indicated that, according to Cloninger’s neurobiological model of temperament, for people experiencing trauma, phobias, depression with anxiety, behaviour related to Avoiding Harm is characteristic. According to Dąbek [42], food is a common regulator of emotions, as well as a way to reduce anxiety. On the other hand, avoidance behaviour is associated with snacking under stress and regarding food as a source of pleasure and mood regulation [42].
PRACTICAL APPLICATION OF RESULTS
Expertise in the care of patients with alcohol dependence shows that the Novelty Seeking and Harm Avoidance temperamental traits, which are mainly expressed in addicts, can make it possible to adapt and target appropriate diagnostic and, consequently, educational or therapeutic methods.
Research has confirmed that the aforementioned traits are not just related to appetitive behaviour directed towards the search for alcohol (alcohol craving, deeper addiction) but also to the search for food (stronger craving) or the relief of pain by taking painkillers.
The temperamental conditioning of appetitive behaviour, such as an increased need to eat, translates into anthropometric parameters, a feature worth taking into account in the treatment of alcohol-dependent individuals.
The present study is particularly relevant because studies of appetitive behaviour in terms of temperamental traits usually focus more on alcohol cravings and less on other behaviours such as nutrition or pharmacotherapy. The presentation of these aspects in the treatment of alcohol dependence complements the holistic view of the patient. For practical reasons of addiction treatment, it becomes essential to identify patients’ temperamental characteristics for a better understanding of their behaviour and also to assess the risk of discontinuation and alcoholic relapse.
LIMITATIONS OF THE STUDY
Besides the positive contributions of the study, the authors note the following limitations:
– the small number of women included in the study,
– the small group of subjects, especially regarding the analysis of the frequency of painkiller use,
– lack of detailed analysis of sources of pain in the study subjects (reasons for taking analgesics),
– lack of use of the standardisation of HS,
– the authors should use more advanced methods for assessing the eating habits of patients.
CONCLUSIONS
In alcohol-dependent individuals, the more expressed Novelty Seeking temperamental trait is associated not only with stronger feelings of alcohol craving and increased hunger but also with a greater appetite. Increased alcohol cravings and hunger may amplify the need for protein-rich and fatty foods.
In alcohol-dependent subjects, the more expressed Harm Avoidance temperamental trait is associated not only with increased craving for alcohol but also with a greater need to relieve pain by taking painkillers and with a higher concentration of serum neuropeptide Y (determined in week 6 of hospitalisation).
In alcohol-dependent individuals, the more expressed Novelty Seeking temperamental trait is associated with impulsive or decreased control of eating and more irritability around hunger.
