Does Reflective Thinking Improve the Ability to Plan in Light and Heavy Drinkers? A Clinical-Experimental Model of Metacognition

Introduction

The main neurocognitive deficits associated with alcohol dependence are impairments in executive functions, short-term and working memory, and attention. This study references the metacognitive model of executive functions proposed by Borkowski and Burke [1] to investigate planning deficits in alcohol-dependent men [2,3] in the context of metacognitive difficulties [4–8]. A broader perspective on planning – including initiation, strategy selection, and correction – appears more appropriate for understanding the nature of these deficits in alcohol-dependent men. The neuropsychological examination was only conducted among men to minimize sample heterogeneity and to specifically focus on functioning in men with alcohol dependence for several reasons: (1) alcohol-related problems occur more frequently among men, who constitute the predominant group in addiction treatment centers; (2) there are known differences between men and women in patterns of brain damage, as well emotional and neurocognitive effects; (3) examining comparable samples of women and healthy individuals would require a longer study period and greater financial resources; and (4) the novel aspect of this research was examination of similarities and differences within a group of men with alcohol use disorder rather than a comparative analysis of typical groups, which has already been extensively described in the literature [9–11].

Metacognition, defined as the monitoring of cognitive processes, is often described as “thinking about thinking” [12] or “cognition about cognition”. Contemporary approaches to metacognition include perspectives that link metacognitive processes to creativity. The creative process allows central components of metacognition to be detected at multiple levels and with varying dynamics; however, these processes require further investigation across diverse clinical groups, including individuals struggling with addiction [13]. Another contemporary perspective on metacognition concerns individuals with deafness, who often experience substantial communication difficulties and social isolation. A novel deep learning model based on micro-electro-mechanical systems (MEMS) and deflection sensors for hand-gesture recognition enables accurate identification of diverse hand movements, which may facilitate communication with the environment, support more controlled learning, and yield a better understanding of others’ reactions [14]. Moreover, the development of metacognitive processes may result not only from improvements in nonverbal communication involving gestures but also from the strengthening of visual perception [15], which may be relevant to research concerning planning skills that rely on visual material. According to another conceptualization, metacognitive processes are related to the ability to assess one’s self-confidence when making decisions [16]. Consequently, a lack of such self-assessment may result in overconfidence, a pattern frequently observed in alcohol-dependent individuals. Paradoxically, speech-related parameters (eg, apnea-hypopnea) may also serve as an indicator of control over the speech production process [17], such as contexts where an individual intentionally withholds information (potentially leading to changes in speech fluency). Furthermore, metacognitive processes may differ between individuals with sleep apnea [18] and those with other somatic diseases, a factor that should be considered in research studies.

Two levels of cognitive structure can be distinguished: meta (related to self-knowledge) and object-oriented (concerning cognitive and affective processes) [19]. Metacognitive disorders are associated with difficulties in learning new abilities, particularly when the learning process is unstructured and influenced by higher impulsivity. Disinhibited, alcohol-dependent individuals often demonstrate such metacognitive deficits [20]. One study of 91 individuals addicted to alcohol, aged 22 to 80 years, showed a negative correlation between metacognitive beliefs and abstinence time. Additionally, differences were observed between groups in early versus full remission, particularly regarding positive and negative metacognitive beliefs (positive metacognitive beliefs about emotional self-regulation and cognitive self-regulation; negative metacognitive beliefs about lack of control and cognitive damage). Those with longer abstinence exhibited worse metacognitive beliefs [21]. Such beliefs may constitute a key problem in recovery and explain persistent difficulties during and after treatment (especially psychotherapy). From a neuropsychological perspective, metacognitive strategies have been incorporated into neurorehabilitation [22]. However, there is a thin distinction between the dominance of metacognition and the specificity of executive functions, suggesting that drinking-related issues should be understood in the context of goal-directed behavior.

Executive functions are defined as intentional, purposeful, and organized behaviors intended to initiate and effectively implement specific actions [23–26]. According to Pąchalska et al [27], 3 components of executive functions can be distinguished: (1) starting, (2) stopping, and (3) shifting. Lezak et al [26] described 4 aspects of executive functions: (1) volition, (2) planning, (3) purposive action, and (4) effective performance. Jodzio [28] proposed 3 concepts of executive functions: (1) a control system, (2) a planning system, and (3) a multifactor model. The present study primarily explored the planning system concept. Planning disorders are not always observed in alcohol-dependent men – they frequently demonstrate effective planning in obtaining alcohol. However, their strategies for broader life functioning are often maladaptive. Other factors, such as post-traumatic stress disorder in adolescence and family history, may increase risks of problem-solving deficits and planning impairment later in life [29]. Additionally, patients with alcohol-related cerebral atrophy, as well as those with symptoms of Korsakoff syndrome or Wernicke encephalopathy, often display executive dysfunction and visuospatial deficits, resulting in visual planning difficulties [30].

The metacognitive model of executive functions is an integrated theory encompassing various aspects of executive functions, such as monitoring, strategy use, and contextual thinking. It also emphasizes that certain thought processes are specific to metacognition [1]. This theory is important for more fully explaining goal-directed behavior in alcohol-dependent individuals. The scheme of the metacognitive model of executive functions includes 3 crucial components: (1) metacognitive knowledge (knowledge about oneself), (2) strategy knowledge, and (3) specific knowledge. Metacognitive knowledge is not clearly defined, but it is specifically focused on metacognition: whether patients demonstrate reflection (socioemotional components) or consideration of how to perform tasks requiring executive function. Strategy knowledge serves as a clear boundary between metacognitive processes and components of executive functions [26,27]. It concerns the selection and correction of strategies during task performance. Specific knowledge refers to everyday knowledge necessary for daily function and goal achievement. All of these components are activated during task performance and may be treated as moderators or mediators of task execution.

Due to the non-standardized nature of this clinical study, an exploratory analysis approach was used, instead of a hypothesis-testing model. The aim of the clinical experiment was to examine visual planning ability and identify metacognition level, considering knowledge, strategy selection and correction, and monitoring capacity, based on the metacognitive model of executive functions proposed by Borkowski and Burke [1]. Two main concepts of executive functions are distinguished: the control system and the planning system [22]. Few published studies have addressed the planning process in individuals with alcohol dependence – most primarily focus on cognitive control. Thus, it is important to investigate this less commonly examined aspect of executive function in alcohol dependence. Three stages of the study were designed, corresponding to the following research questions: (1) Are there differences among patients in metacognitive level and executive capacity as measured by visual planning tasks? (2) Does visual planning ability in alcohol-dependent individuals change after metacognitive activation?

Material and Methods

PARTICIPANTS:

One hundred twenty-six patients were examined; however, 3 did not complete the neuropsychological examination, and the examination of 1 patient was discontinued due to severe brain damage observed on magnetic resonance imaging. Therefore, analyses were conducted in 122 men. All patients were diagnosed with alcohol dependence syndrome (International Classification of Diseases [ICD]-10 F10.2) and alcohol use disorder (using Diagnostic and Statistical Manual of Mental Disorders [DSM]-5 classification). The target participant age range was 26 to 59 years. The project initially included older individuals, but the age range was revised when younger alcohol-dependent men could be recruited at several centers. Although 3 older patients were subsequently excluded from the main project, the analysis in this article includes the entire examined group of patients with alcohol use disorder, aged 26 to 66 years (median [Med]=44, mean [M]=44.05, standard deviation [SD]=8.59), with various years of education (Med=12, M=11.80, SD=2.03, Min=8, Max=17, IQR=11–12). The addition of older patients (3 individuals aged >59 years) did not strongly influence the results. A structured interview was conducted at the recruitment stage, in compliance with ICD-10 and DSM-5 diagnostic criteria, to identify eligible patients. The inclusion criteria were: (1) diagnosis of alcohol dependence (ICD-10 and/or DSM-5 criteria) and (2) age between 26 and 59 years (main group). The exclusion criteria were: (1) frequent ongoing use of other psychoactive substances, (2) regular use of neuroleptics due to other mental or behavioral disorders, (3) somatic diseases not resulting from alcohol addiction, and (4) post-traumatic brain dysfunction (if undocumented or insufficiently explained by participants). Patients with severe neurological disorders (eg, dementia, stroke, or other neurodegenerative diseases) and/or severe psychiatric disorders (eg, schizophrenia, major depression, or other disorders) also were excluded from the study [31]. Detailed participant data are presented in Table 1.

One hundred nine patients (89.3%) were right-handed, 5 patients (4.1%) were ambidextrous, and 8 patients (6.6%) were left-handed. Because the vast majority of participants were right-handed (89%) and only a small proportion were left-handed or ambidextrous, the effects of handedness on cognitive processes and alcohol consumption intensity were not determined. Table 2 shows the patient profile according to alcohol consumption level.

The screening was carried out via coordination with psychologists, psychiatrists, and nurses working at the treatment and support centers. Neuropsychological examinations were conducted by the lead investigator. All patients provided written consent to participate in the study, and the study protocol was approved by the local Ethics Committee (Faculty of Psychology and Cognitive Science, Adam Mickiewicz University, approval number 8/03/07/2018).

METHODOLOGY:

All patients were assessed as part of a broader screening project to collect demographic and alcohol consumption data, and the research described in this article is a key component of that project. Alcohol use disorder severity was measured using the Polish adaptation [32] of the Michigan Alcoholism Screening Test (MAST) [33]. The MAST consists of 24 items addressing alcohol-related problems, such as the need for hospitalization due to alcohol dysfunction. Participants respond either “yes” or “no”; each response is assigned a value of 1, 2, or 5 points for diagnostic purposes. The final score is the sum of all response values; a score higher than 4 indicates the presence of alcohol dependence [34]. A previous report showed that the reliability of the MAST, measured using Cronbach’s alpha in a sample of 256 alcohol-dependent individuals (71 women and 185 men), was 0.822 [35]. The tool demonstrates good psychometric properties [36,37]. No cutoff threshold was used to distinguish light and heavy drinkers because test items are not scored in a consistent manner; a score of 5 may indicate mild alcohol-related problems in some cases and considerably greater severity in other cases. Generally, higher MAST scores reflect greater alcohol-related problems. The Statistical Analysis section presents an alternative approach to distinguish between light and heavy drinkers in the present study. Additional data were collected through neuropsychological research.

CLINICAL EXPERIMENT TO MEASURE EXECUTIVE FUNCTIONS:

This study explored neuropsychological symptoms in men with alcohol use disorder, particularly concerning executive functions and metacognitive processes. The tools used are described below.

The Porteus Maze Test [38] measures psychological planning capacity (nonverbal domain) and foresight in children, adolescents, and adults. The test consists of a series of paper forms that present drawings of mazes with increasing complexity. An individual taking the test is required to trace a path through the maze within a time limit of 15 to 60 minutes. The individual must avoid blind alleys and dead ends; no backtracking is allowed. Only 2 versions of the test (Porteus Maze Test Adult I – No. 37055O Porteus Test Vineland Revision and Porteus Maze Test Adult II – No. 37055R Porteus Test Vineland Revision) were used in this study under the License Agreement of August 28, 2018, and Supplement No. 1 to the License Agreement of July 31, 2019, from Stoelting Co. (https://www.stoeltingco.com/). Participants were allowed 1 test attempt each at the pre-test stage (Porteus Maze Test Adult I) and post-test stage (Porteus Maze Test Adult II).

All patient responses were recorded as stated in the study protocol. After all experimental steps had been completed, patient mood (normal/lowered/elevated) and thinking dynamics were assessed, including normal/accelerated/slow thinking, as well as inertia and/or lability of thought processes. The stages of the clinical experiment are presented in Figure 1.

STATISTICAL ANALYSIS:

The statistical analysis included: (1) correlation analysis for quantitative variables without a normal distribution (Spearman’s rho coefficient) and for nominal variables (Cramér’s V or Kendall’s τ coefficient, depending on variable type), and (2) multivariate mixed-model MANCOVA (2-way repeated-measures analysis of covariance with cross-participant factors) to verify visual planning abilities before and after metacognitive activation in all groups of patients. The dependent variable was visual planning ability, whereas the independent variables were related to 4 cross-participant factors: (I) visual planning ability group, (II) activation – monitoring and metacognitive knowledge, (III) activation – strategy knowledge, and (IV) activation – specific knowledge. Covariates were age and education (main control variables), as well as alcohol abuse duration and alcohol use disorder severity (main alcohol-related variables). In the first step, the interaction effects of all activation factors on visual planning were measured. In the second step, the interaction effects of group membership, the role of activation factors, and visual planning were examined. Consistent with the research problem, only a limited set of interaction effects was considered. Finally, differences were assessed between groups of patients with proper planning and those with planning difficulties, particularly concerning activation of specific knowledge. Table 3 presents the statistical analyses used in the clinical experiment.

Results

Correlations of Visual Planning, Estimated Metacognitive Level, and Executive Capacities

VISUAL PLANNING NO. 1 (NUMBER OF ERRORS):

First, correlations of visual planning with quantitative variables were examined using Spearman’s rho. Significant negative correlations were detected between the number of errors in visual planning and both visuospatial function (rho(122)=−0.307, P<0.001) and general Addenbrooke’s Cognitive Examination (ACE-III) score (rho(122)=−0.221, P=0.014). Thus, a greater number of errors in visual planning was associated with worse basic cognitive function, particularly due to difficulties in visuospatial processing. There is a very weak statistically significant correlation between age and number of errors in visual planning (rho(122)=0.178, P=0.05). No statistically significant correlations (P>0.05) were observed between visual planning and level of education, or alcohol consumption.

Correlations of visual planning with qualitative variables were examined using Kendall’s τ coefficient. There were no statistically significant correlations (P>0.05) between visual planning and marital status, education, other mental disorders and microinjuries, somatic illnesses, or drinking history.

ESTIMATED METACOGNITIVE LEVEL AND ESTIMATED EXECUTIVE CAPACITIES: Correlations of nominal variables with estimated metacognitive level and estimated executive capacities were explored using Cramér’s V coefficient. Statistically significant correlations with estimated metacognitive level and estimated executive capacities were observed for selected qualitative control variables. The results are presented in Figure 2.

Statistically significant correlations were found between: (A) patients’ family history of alcohol problems and estimated metacognitive level measured by question 1 (V(111)=0.247, P=0.034), (B) marital status and estimated metacognitive level measured by question 2 (V(122)=0.222, P=0.049), and (C) level of education and estimated executive functions measured by question 1 (V(122)=0.262, P=0.002). There was no statistically significant correlation between level of education and estimated executive functions measured by question 2 (V(122)=0.137, P=0.600), nor between either estimated metacognitive level or executive capacities (functions) and other qualitative variables (P>0.05).

There also were no correlations between visual planning and any aspects of estimated metacognitive level and executive capacities (P>0.05). Several correlations were observed between some aspects of estimated metacognitive level and estimated executive capacities (functions). Details are presented in Figure 3.

Statistically significant correlations were found between: (A) estimated metacognitive level measured by question 1 and estimated metacognitive level measured by question 2 (V(122)=0.333, P<0.001), (B) estimated executive capacities measured by question 1 and estimated executive capacities measured by question 2 (V(122)=0.512, P<0.001), (C) estimated metacognitive level measured by question 2 and estimated executive capacities measured by question 1 (V(122)=0.269, P<0.001), (D) estimated metacognitive level measured by question 1 and estimated executive capacities measured by question 1 (V(122)=0.286, P<0.001). No other statistically significant correlations were observed.

GROUPING BASED ON VISUAL PLANNING ABILITY: The Porteus Maze Test is a clinical-experimental neuropsychological tool, rather than a psychometric model [28]. Thus, precise thresholds for normal and impaired levels of planning ability cannot be defined. Nevertheless, based on previous planning process research involving labyrinth tasks [39], performance was classified according to the number of errors using a cutoff threshold of 2, where a score below 2 indicates normal planning and a score above 2 indicates impaired planning ability. In the present study, visual planning processes (Porteus Maze Test results) were classified into 2 categories: proper planning and planning difficulties. To examine the influence of metacognitive processes on visual planning ability, 2 groups were identified: (1) patients with PROPER PLANNING ABILITY (N=95; <2 pre-test errors) and (2) patients with PLANNING DIFFICULTIES (N=27; >2 pre-test errors). There were no statistically significant (P>0.05) differences between these 2 groups in any alcohol consumption variables; there were statistically significant differences related to age [F(1,120)=4.54, P=0.035, η2=0.12], basic cognitive function measured by ACE-III [U=865, P=0.010], Mini-Addenbrooke’s Cognitive Examination (M-ACE) [U=945, P=0.036], phonemic fluency [F(1,120)=4.97, P=0.028, η2=0.13], semantic fluency [F(1,120)=8.94, P=0.003, η2=0.17], and visuospatial function [U=762, P<0.001]. Patients in the PROPER PLANNING group, who also were younger alcohol-dependent men, obtained higher scores on these cognitive tasks.

CORRELATIONS BETWEEN RESPONSES IN THE FIRST STEP OF ACTIVATION AND PLANNING ABILITIES:

Step 1 of activation was conducted in patients who likely experienced some visual planning difficulties in the pre-test, as suggested by their comments during the test. It also was conducted in patients without planning difficulties. Statistically significant correlations were found between: (1) visual planning ability group and responses to question 1 in Step 1 of activation (χ²=15.7, P<0.001) and (2) visual planning ability group and responses to question 2 in Step 1 of activation (χ²=19.5, P<0.001).

RESPONSES ABOUT USING METACOGNITIVE KNOWLEDGE AND ITS STRATEGIES IN VISUAL PLANNING WERE EXAMINED IN RELATION TO ALCOHOL CONSUMPTION VARIABLES: No statistically significant differences were detected concerning responses to Monitoring & Metacognitive Knowledge and Strategy Knowledge questions with regard to alcohol abuse duration (P>0.05), abstinence duration (P>0.05), or daily alcohol consumption (P>0.05). However, a statistically significant difference was observed in responses to Monitoring & Metacognitive Knowledge questions according to alcohol use disorder severity [F(1,120)=4.017, P=0.047, η2=0.032]. There was no statistically significant difference in responses to Strategy Knowledge according to alcohol use disorder severity (P>0.05). The results are presented in Figures 4 and 5.

LIGHT VERSUS HEAVY DRINKERS: DIFFERENCES IN RESPONSES CONCERNING THE USE OF SPECIFIC KNOWLEDGE TO IMPROVE VISUAL PLANNING ABILITIES DURING ACTIVATION PROCESSES: Given that statistically significant differences were evident among patients in terms of their approach to specific knowledge, the roles of age, education, and alcohol consumption variables were examined. No statistically significant differences were observed between patients with ambiguous responses (more reflective) and those with unambiguous responses in terms of age, education, basic cognitive function, or multiple aspects of alcohol consumption (P>0.05), excluding alcohol use disorder severity [F(1,120)=5.810, P=0.017, η2=0.046] and abstinence duration [U=1432, P=0.047]. Patients who confirmed that knowledge about their skills helped to improve visual planning were characterized by higher alcohol use disorder severity (more symptoms on the MAST scale) [N=68; M=38.13; SD=13.849] compared with patients who answered less clearly (denied or confirmed with more explanation) [N=54; M=32.26; SD=12.733]. Additionally, longer abstinence duration was observed in patients who confirmed that knowledge about their skills helped to improve visual planning [N=67; M=47.61; SD=25.093] compared with patients who answered less clearly (denied or confirmed with more explanation) [N=54; M=40.20; SD=31.734]. Intriguingly, longer abstinence was sometimes observed among heavy drinkers. Partial results are presented in Figure 6.

EFFECT OF METACOGNITIVE ACTIVATION ON VISUAL PLANNING ABILITIES IN ALL PATIENTS:

A multivariate mixed-model MANCOVA was conducted to address the research problem.

First, the design included 3 cross-participant factors (activation processes, added gradually), 1 within-participant factor (visual planning abilities before and after activation), and covariates (age, education, alcohol abuse duration, and alcohol use disorder severity). The dependent variable was the average result in terms of executive functions (calculated as the inverse of the number of errors in the visual planning task; when zero errors occurred, a value of “2” was assigned proportionally). The Wilks’ Lambda test was performed, and Bonferroni correction was applied. Only effects relevant to the research problem were observed, as follows: correlation between Activation (Monitoring & Metacognitive Knowledge) and visual planning abilities [F(1,115)=0.285, P=0.594, η2=0.002], Activation (Strategy Knowledge) and visual planning abilities [F(1,115)=0.703, P=0.404, η2=0.006], and Activation (Specific Knowledge) and visual planning abilities [F(1,115)=6.612, P=0.011, η2=0.054]. Pairwise comparisons of patients with different responses in the Activation (Specific Knowledge) condition showed that patients who answered with greater reflection (N=54; responses “no,” without a clear answer, or “yes” with more explanation) demonstrated improved visual planning in the second measurement (P<0.05) relative to patients who answered with less reflection but gave a clear, unambiguous answer (N=68; patients who confirmed that knowledge about their skills is helpful) (P>0.05). Covariates showed no interaction effect with regard to the dependent variables. The results are presented in Figures 4–6.

EFFECT OF METACOGNITIVE ACTIVATION ON VISUAL PLANNING ABILITY GROUP:

Second, the design included 4 cross-participant factors (group membership and 3 activation processes added gradually), 1 within-participant factor (visual planning abilities before and after activation), and covariates (age, education, alcohol abuse duration, and alcohol use disorder severity). The dependent variable was the average result in terms of executive capacities (calculated as the inverse of the number of errors in the visual planning task; when zero errors occurred, a value of “2” was assigned proportionally). The Wilks’ Lambda test was performed, and Bonferroni correction was applied. The analysis was conducted separately, as follows: (1) interaction effect of group membership (group with planning difficulties and group with proper planning) and visual planning [F(1,115)=39.315, P<0.001, η²=0.255] (the interaction effect of group membership in the analysis with all activation processes also was statistically significant for group, P<0.001), (2) interaction effect of Activation (Monitoring & Metacognitive Knowledge), group membership (group with planning difficulties and group with proper planning), and visual planning abilities [F(1,113)=0.000, P=0.994, η²=0.000], (3) interaction effect of Activation (Strategy Knowledge), group membership (group with planning difficulties and group with proper planning), and visual planning abilities [F(1,113)=0.080, P=0.778, η²=0.001], and (4) interaction effect of Activation (Specific Knowledge), group membership (group with planning difficulties and group with proper planning), and visual planning abilities [F(1,113)=1.679, P=0.198, η²=0.015].

Pairwise comparisons in the group of patients with planning difficulties showed that those who made more errors in visual planning before activation significantly improved their visual planning abilities (P<0.05). However, activation type had no impact on this effect (P>0.05). In contrast, pairwise comparisons in the group of patients with proper planning showed that activation was statistically significant only in the domains of strategy knowledge and specific knowledge (P<0.05); intriguingly, their performance in the second visual planning task worsened after activation (P<0.05). Additionally, responses related to specific knowledge significantly differed among patients with proper planning abilities in the second measurement of visual planning (P<0.05).

With regard to covariates, only an interaction effect of age on the dependent variable (visual planning ability) was observed (P<0.05). The MANCOVA results are presented in Figures 4–8. Finally, estimated metacognitive level and estimated executive capacities did not significantly influence the activation process (P>0.05).

Discussion

Correlations of Visual Planning, Estimated Metacognitive Level, and Executive Capacities

IS THERE A DIFFERENCE AMONG PATIENTS IN METACOGNITIVE LEVEL AND EXECUTIVE CAPACITY? FROM THEORETICAL MODEL TO RESEARCH RESULTS:

The present results imply differences among patients with alcohol use disorder in terms of metacognitive level and executive capacities, although these differences are not directly dependent on alcohol consumption. In this study, the metacognitive level in alcohol-dependent men was assessed before the visual planning task to encourage a more reflective approach among included patients. This approach represents a novel methodological step aimed at linking task performance with reflective thinking. The results highlighted the role of family history of alcohol use disorder. Patients with such a history were more likely to consider the task moderate (rather than easy) relative to patients without alcohol problems in their families, who tended to consider it easy. A possible explanation may involve lower self-criticism in alcohol-dependent individuals – often described as a symptom of anosognosia [40] – although it is challenging to identify during clinical diagnosis [41]. Additionally, patients who were single (unmarried [ie, never married], divorced, or widowed) more frequently reported that they would be able to cope with the task before attempting it. This perspective may result from the need to manage most everyday tasks independently, without support from loved ones. The above observations partially confirm the hypothesis that differences are present among patients in terms of metacognitive level, but only regarding family history of alcohol problems and marital status.

Estimated executive capacity after the first visual planning task appeared to vary depending on patients’ education level. Nevertheless, relatively few patients had either only basic or higher education. One notable implication of these results is unsurprising: patients with higher education reported that the task was not difficult, and they tended to provide clear and concise explanations. Similar patterns may be expected in other clinical groups, such as patients with nonalcoholic fatty liver disease [42]. This expectation may require verification in a control group, although – as intended – the visual planning task used here would likely be too easy for healthy individuals.

Intriguing results were observed when comparing patient responses before and after the visual planning task. Patients who initially described the task as easy or moderate generally confirmed afterward that they were able to cope with it, in contrast to those who had anticipated difficulties in this respect. A clear distribution of responses was observed along this dimension. Most patients who initially rated the task as easy, moderate, or difficult gave the same evaluation after performing it; some initially rated the task as moderate and later reported that it had been relatively easy. Similarly, patients who initially expected that they could cope with the task often confirmed (after completion) that it had been easy. Patients who reported that the task had been easy tended to provide more reflective responses and greater explanation, often elaborating on whether the task was performed quickly and without mistakes. These findings suggest that the act of evaluating the visual planning task is an initial step toward more reflective thinking in patients who otherwise may experience difficulties with it, as observed in clinical practice [43] and alcohol-related research [44].

In summary, these results provide insights concerning the role of metacognition in alcohol use disorder, which may be understood as emotional insight and awareness of social functions, reflected in the consistency of patient responses before and after the first visual planning task. Le Berre [20] emphasized the importance of incorporating metacognitive abilities into neuropsychological rehabilitation for alcohol-dependent patients, given that a substantial proportion of such patients show deficits in emotional processing and social cognition. These deficits can lead to impaired insight, expressed as weak self-criticism and a less reflective approach to planning basic daily activities and long-term goals [45]. Estimations of metacognitive level and executive capacity may serve as an important extension of the metacognitive model of executive functions proposed by Borkowski and Burke [1], in which these processes are defined as thinking about strategies, without strict or unambiguous definitions. This definition is not a weakness of the model, but rather a strength – it allows empirical verification.

DOES THE EFFICIENCY OF EXECUTIVE FUNCTION CHANGE UNDER THE INFLUENCE OF METACOGNITIVE ACTIVATION LEVEL? FROM MODEL TO RESEARCH:

Based on the theoretical model [1], thinking process activation involves 3 kinds of knowledge: metacognitive, strategy, and specific. In the present study, it was assumed that responses to 3 corresponding activation questions would be sufficient and meaningful to demonstrate either a reflective or impulsive approach to the process of planning. This assumption allowed verification of a potential improvement in executing the second part of the visual planning task, both in the full sample of alcohol-dependent men and in subgroups. Patient narratives revealed 2 main categories of responses: reflective and impulsive. These responses were given to the 3 questions designed to activate more detailed thinking. This distinction is important because patients with long-term alcohol abuse often show highly impulsive reactions [46]. Notably, reflective thinking does not occur accidentally. Although the concept of reflective thinking remains somewhat elusive, many studies have attempted to capture its essence. The present study aimed to stimulate reflective thinking among patients in the context of performing a task that involved a component of executive functions – planning. The reflection process has been described as “an internal dialogue between oneself as a subject and an object (or object processes)”; moreover, it is “a highly individual and internal process resembling a dialogue with oneself, largely inaccessible to others” [47]. Thus, in clinical practice and research, it can be assumed that more detailed and explanatory responses are indicators of more reflective thinking, whereas impulsive answers are typically quick and unambiguous.

In this clinical experiment, response correctness was irrelevant. The design was adapted to the clinical group studied – patients with alcohol dependence – and therefore a control group was not required. Importantly, reflective thinking among alcohol-dependent men has not been fully examined thus far. Research has often focused on impulsivity, and evidence suggests that all impulsivity features (trait, motor, and choice) are associated with alcohol use disorder relapse [48].

ALCOHOL-DEPENDENT MEN WITH HIGHER IMPULSIVITY: HEAVY DRINKERS?: Heavy drinkers are frequently discussed in the literature, but their functional profile is not fully understood. This profile may depend on the extent of structural changes in the brain and strategies for coping with alcohol use disorder. One study suggested that cognitive control is preserved among young adults engaging in heavy drinking [49]. However, heavy drinkers with Wernicke encephalopathy or Korsakoff syndrome show cognitive decline, accompanied by typical structural atrophy of the brain, especially in the frontal lobes, cerebellum, and subcortical structures [50]. Heavy drinkers also exhibit higher alcohol craving after cue exposure compared with occasional drinkers, as noted observed in a virtual bar experiment [51]. Other research yielded 3 important observations: (1) a prominent role of family history of alcohol abuse among heavy drinkers compared with low-risk and high-risk social drinkers, (2) a propensity of heavy drinkers to engage in more binge-drinking days, and (3) consumption of larger quantities of alcohol among heavy drinkers, as measured by total drinks in a 90-day Timeline Followback [52]. In a more recent experiment, alcohol-dependent individuals were categorized as light or heavy drinkers, and the results showed that heavy drinkers were more susceptible to inhibition failure in situations related to alcohol consumption [53]. Similarly, another study explored heavy drinkers’ and non-drinkers’ intentions and behavior based on their similarity to prototypical heavy drinkers and non-drinkers using Alcohol Use Disorders Identification Test – Consumption (AUDIT-C) in young adults. In that study, both intentions and willingness mediated the relationship between perceptions of prototypical heavy and non-drinkers and AUDIT-C scores [54].

There is evidence that heavy drinkers make more inhibitory errors than light or non-drinkers, indicating inhibition deficit and executive dysfunction [55]. Sex differences have also been observed. Female heavy drinkers show greater cognitive control problems [56]; the present findings partially confirmed such monitoring difficulty. This behavioral challenge may explain why many heavy drinkers experience numerous negative effects on health and well-being [57]. Other evidence suggests that heavy drinkers are more impulsive or may have “more experience in exerting cognitive control over attentional biases” [58], which could explain why their responses tend to be quicker and clearer (similar to specific knowledge responses). In a particularly interesting study, Rowicka [59] examined only individuals with alcohol problems (without a control group, as in the present study) and identified 4 categories – current low risk (group 1), current dependent (group 2), former heavy user/current low risk (group 3), and former dependent/current low risk (group 4) – for exploration of drinking motives. The results showed that heavy drinkers did not report motives distinct from those of other groups. Similarly, 4 distinct profiles of individuals with alcohol problems were identified; high-functioning occasional heavy drinkers paradoxically displayed lower alcohol dependence severity relative to high-functioning infrequent non-heavy drinkers [60]. The overall profile of heavy drinkers can also be described in terms of attentional bias: heavy drinkers demonstrate significantly greater attentional bias than moderate drinkers [61] and more frequently show a preference for alcohol-related choices compared with light drinkers [62].

ALCOHOL-DEPENDENT MEN WITH MORE REFLECTIVE THINKING: LIGHT DRINKERS?: The present results indicate that the activation of more reflective thinking was effective, especially in terms of the metacognitive model of executive functions proposed here. Alcohol-dependent men showed improvement in visual planning abilities after reflection that involved incorporating specific knowledge into the planning process; this effect was limited to patients who responded more reflectively (likely indicating lower impulsivity) and who were simultaneously light drinkers. These observations highlight a difference in reflective thinking compared with patients who unambiguously confirmed that specific knowledge about their skills is helpful. Clear responses and unequivocal confirmation may often reflect greater impulsivity (or heightened arousal), meaning that rapid positive responses such as “yes” or “I agree” could be purely spontaneous and poorly controlled. Heavy drinkers typically display higher impulsivity and reduced self-control. The current results partially support previous research suggesting stronger arousal in heavy drinkers and occasional improvements in cognitive function [63]. Activation of metacognitive thinking in heavy drinkers may require more advanced interventions due to the motivational mechanism of strong dependence [64]. The current results also suggest that light drinkers exhibit lower alcohol use disorder severity, fewer alcohol dependence symptoms, and less serious consequences (eg, poisoning treatment or legal problems). Light drinkers more often gave ambiguous responses, without clear confirmation (but with added explanation). In another study, light drinkers had a lower average score on the Alcohol Use Disorders Identification Test (AUDIT) screening tool (3.0 points) compared with heavy drinkers (11.4 points). The groups also differed in terms of sleep and circadian variables: light drinkers reported shorter sleep duration but exhibited fewer signs of insomnia and no elevated risk [65]. A particularly notable study concerned alcohol use and inhibitory control. Although findings were mixed regarding differences between heavy drinkers and light drinkers, exposure to alcohol-related image stimuli disrupted inhibitory control in heavy drinkers, but not in light drinkers [53]. Additionally, light drinkers exhibit a lower risk of cancer mortality relative to non-drinkers and heavy drinkers [66]. Among Brazilian adults, light drinking was defined as daily ethanol intake of 1 to 12.9 g per day [67], whereas a study of middle-aged men in the United States used a definition of 5 to 14 drinks consumed over the previous 14 days [68]. Distinctions between light drinkers and heavy drinkers have also been observed in student populations [69], which differ from alcohol-dependent individuals such as the present study cohort.

ALCOHOL-DEPENDENT MEN WITH PRESERVED METACOGNITION AND COGNITIVE DIFFICULTIES: The activation of metacognitive beliefs improved visual planning among both light and heavy drinkers who initially showed visual planning difficulties. This effect was not observed in patients with proper planning abilities. Intriguingly, activation type did not have a statistically significant effect, which is partially consistent with previous findings. A particularly informative study addressed 3 aspects of function – attentional bias, craving, and metacognition – in a small sample of alcohol-dependent individuals. The key findings indicated significant interactions of all metacognitive beliefs (negative beliefs about thoughts concerning uncontrollability and danger, cognitive confidence, and beliefs about the need to control thoughts) with responses to alcohol-related image stimuli [70]; other results were less clear. In the present study, improved visual planning abilities were observed after activation of metacognitive beliefs (measured through responses to questions regarding monitoring and the use of metacognitive knowledge, strategy knowledge, and specific knowledge) in patients with planning difficulties who also displayed basic cognitive disorders, especially in terms of visuospatial function. These findings suggest that such patients experience brain dysfunction in posterior parts of the brain, rather than exclusively showing severe prefrontal atrophy [71–73].

ALCOHOL-DEPENDENT MEN WITH PRESERVED BASIC COGNITION AND METACOGNITIVE DIFFICULTIES: This study showed that the visual planning task was relatively easy; its performance was not correlated with alcohol consumption, age, or education for the majority of patients. Most alcohol-dependent men with proper planning abilities (heavy and light drinkers) successfully completed the first visual planning task. These results can be explained from the perspective of metacognition (ie, reflective thinking). When a task involving executive function is easy, engagement in additional reflection may lead to overthinking and slower task performance [74]. In the present study, certain aspects of activating metacognitive beliefs (related to strategies and specific knowledge) led to worse visual planning ability in both heavy and light drinkers who had shown no difficulties in the first planning task. This effect may be interpreted as a consequence of excessive reflective thinking (positive functional aspect) or overthinking (negative functional aspect). In the literature, there is little information about overthinking among alcohol-dependent individuals, possibly due to the lack of such patterns in those individuals. However, some findings indicate that repetitive thinking can predict alcohol use disorder severity in individuals with strong metacognitive beliefs, such as negative beliefs about uncontrollability and danger. Additionally, repetitive negative thinking has been shown to predict alcohol use disorder severity in men with high metacognitive beliefs about the need to control thoughts and with high positive metacognitive beliefs concerning emotional self-regulation [75]. Metacognitive difficulties may also be explained by mechanisms of neuroplasticity [76] and the microgenetic approach, which provide insight into the onset of metacognitive symptoms [77,78]. These impairments should be addressed through metacognitive training aimed at reducing negative symptoms [79] or psychosis-related processes [80], which may also emerge in heavy drinkers.

A notable limitation of this study was the heterogeneous participant profile, characterized by a wide age range, varied abstinence duration, and diverse clinical statuses. However, such heterogeneity is broadly typical of alcohol-dependent populations; creating homogeneous groups in terms of age, abstinence duration, and clinical status could introduce serious bias. Therefore, adjustments were attempted for these variables. Another important limitation concerned the procedure of the clinical experiment and the measurement of planning processes. For example, instead of the Porteus Maze Test, other widely used instruments (eg, Tower of London Test) could have been used. Although the experiment was based on a validated model of metacognitive processes and executive functions, it has not been updated or tested in diverse clinical groups thus far. Further studies are needed to replicate these findings in other clinical groups.

Conclusions

Metacognitive skills, or more reflective thinking, among alcohol-dependent individuals were evident in their responses before and after completion of 2 simple planning tasks. Differences were observed in the responses of light drinkers relative to those of individuals with more severe alcohol-related problems (ie, heavy drinkers). Future research should further distinguish between light and heavy drinkers, assess other skills related to planning, and examine the manner in which alcohol-dependent individuals express themselves, given that verbal expression can reflect biopsychosocial function, an important consideration when developing treatment programs for this population.