While they have been far less explored than cerebral and cognitive alterations in alcohol-dependence, emotional and interpersonal ones are omnipresent in alcohol-dependence and highly involved in relapse 
. Unfortunately, little is known about the biased representations underlying these affective and social difficulties. As maladaptive beliefs about the self in social contexts influence emotional and social processes in other psychopathological states 
, the present study investigated social self-beliefs in ADS, and their links with interpersonal problems. A main result was that ADS exhibit significantly higher scores than CS for SBSA total and for every subscale, which constitutes the first description of a general bias for maladaptive self-beliefs related to social contexts in addictive states. Nevertheless, similarly biased self-beliefs have been found in general and social anxiety 
, and these psychopathological states are frequently associated with alcohol-dependence, as confirmed by the comorbidities observed in the present sample. These comorbidities were thus taken into account in a second analysis to determine their own impact on maladaptive self-beliefs, aside from the direct effect of alcohol-dependence.
As illustrated in , group differences on SBSA global score, conditional beliefs and unconditional beliefs disappeared when these comorbidities were introduced as covariates, but the ADS still presented significantly higher scores for the high standards subscale. Hence, a central observation of the present study is that alcohol-dependence, beyond anxious and depressive comorbidities, is associated with excessively high standards in social contexts. ADS thus seem to particularly overvalue requested behavior in interpersonal context (e.g., “I have to appear intelligent and witty”) and their obligation to constantly have a perfect social behavior (e.g., “I must get everyone's approval”). In other words, they overestimate the level of interpersonal performance needed to obtain the desired positive social outcome. The potential reactions of others if this social performance is not optimal (conditional beliefs) and the general negative opinions of others concerning subject's social behaviors (unconditional beliefs) are also overrated, but these biased cognitions seem to be mostly a result of the comorbid conditions that these patients frequently present with. Nevertheless, we acknowledge that covariate analyses do not allow us to clearly determine what part of variance related to comorbidities is removed 
, and thus in interpreting our covariates analyses we must keep this in mind.
However, it has been recently shown that social anxiety is associated with excessive high standards, conditional beliefs and unconditional beliefs 
. The present results together with earlier ones thus suggest that, while maladaptive self-beliefs are present in a wide range of psychopathological states (including bulimia nervosa 
and psychosis 
) and are thus transdiagnostic, the three categories of self-beliefs could be distinctly distorted in different psychopathological states. Finally, these exaggerated high standards are strongly associated with the social problems encountered in real-life situations. Indeed, while interpersonal problems were not linked with SBSA total score or conditional/unconditional beliefs subscales, high standards scores were highly correlated in the ADS with IIP. No causal inference can be drawn from these correlational results, but they nevertheless confirm that biased cognition in social context is intimately connected with the actual behaviors and performance in these contexts.
These results have important implications. At fundamental level, they offer the first insights into the biased psychological representations involved in the affective and social difficulties described in alcohol-dependence. ADS present massive impairments in emotional and interpersonal situations 
, and this pathology is marked by strong social stigma 
as well as higher sensitivity to social rejection 
. The present results showing the links between maladaptive self-beliefs and social deficits could thus constitute a first step towards a thorough exploration of the possible pathways linking biased psychological representations, social alterations and alcohol-dependence. First, an individual who holds excessively high social standards may frequently have difficulties reaching those unrealistic standards in social contexts, and the resulting distress could be temporally reduced by increasing the use of alcohol as a self-medication. Second, individuals with excessively high social standards may believe that alcohol enhances their social abilities, and may thus increase their alcohol consumption to meet their high social standards. Third, problematic alcohol use and related interpersonal difficulties may already be present for an individual. Excessively high social standards may then develop to guide behaviour so as to attempt to avoid interpersonal conflict, and allow alcohol use to continue. The present results are consistent with these pathways but the specific temporal ordering of the variables have to be examined further in future studies. More globally for psychopathology, our results show a significant correlation between maladaptive self-beliefs and the intensity of interpersonal problems in everyday life. This observation should be confirmed and extended among different psychopathological populations, but it offers the first confirmation of the association between maladaptive self-beliefs and self-reported difficulties with social behaviors.
At a clinical level, these results suggest that maladaptive self-beliefs negatively influence social integration in alcohol-dependence and could favour the maintenance of alcohol consumption by reinforcing the vicious circle between interpersonal problems, social isolation and alcohol consumption. Therapeutic interventions reducing this self-beliefs bias might thus interfere with the vicious circle and reduce alcohol consumption or relapse risk. Interventions focusing on maladaptive self-beliefs have been tested in social anxiety 
. Specifically, “widening the bandwidth” exercises 
aim at questioning the high standards during social performance and showing that the range of adapted social behaviors is larger than those allowed by these high standards. Proposing these brief exercises during the rehabilitation process might lower the exaggerated high standards present in ADS, and improve social functioning after detoxification.
Several limitations have to be underlined. First, while the results observed are highly significant and based on strong statistical power, the sample size was quite limited. Larger studies should be conducted, particularly to explore the heterogeneity of ADS. Second, inappropriate social behaviors were assessed using self-report. Future studies could usefully examine the associations between self-beliefs and behavioral indices of social behaviors, measured by an experimental approach complementing our questionnaire-based approach (e.g., multimodal assessment of social behaviors 
). Third, this first exploration in alcohol-dependence should be followed up by explorations of maladaptive self-beliefs in other psychopathological states in which emotional and interpersonal alterations are highly present (e.g., schizophrenia and autism), in order to confirm whether distorted self-beliefs in a social context constitute a transdiagnostic process. More specifically, exploring these self-beliefs in other addictive states would give important insights on the common nature of different addictions. Moreover, the links between these maladaptive self-beliefs and core psychological processes in alcohol-dependence (e.g., rumination, psychological craving, avoidance of social interactions, lack of self-forgiveness 
) should be further explored to better understand the mediating variables linking the influence of these biased cognitions on social behaviors. Finally, the proposal that these exaggerated high standards hamper treatment seeking should also be tested, as increasing the proportion of patients included in detoxification process is a crucial public health goal in alcohol-dependence 
To conclude, this study sought to explore for the first time whether self-beliefs related to social interactions are characteristics of ADS. Our main prediction that alcohol-dependence would be associated with maladaptive self-beliefs was strongly supported, as shown by significantly higher scores in ADS than CS for SBSA global score and subscales. Critically, when the comorbidities (depression, anxiety and social anxiety) were taken into account using covariate analyses, the only remaining significant bias among ADS was the presence of exaggerated high standards during social interactions. Independently from common comorbidities, alcohol-dependence is thus linked with exaggerated high standards in interpersonal contexts, which could contribute to the vicious circle between alcohol consumption, interpersonal problems and social isolation 
. Indeed, significant correlations were also found between excessive high standards and interpersonal problems, suggesting a direct link between maladaptive self-beliefs and actual social difficulties. While these results should be confirmed and extended in future studies, they nonetheless provide the first insights into the influence of underlying cognitive biases on interpersonal problems in addictive states, and thus open exciting possibilities for future investigations in this area.