آلکسی تایمیا، یک عامل ترکیبی برای علائم غذا خوردن و اجتناب اجتماعی در بیماری بی اشتهایی عصبی

Socio-affective difficulties, in particular difficulties in representing, communicating and feeling emotions, may play a critical role in anorexia nervosa (AN). The aim of this longitudinal study was to explore the links between alexithymia and two types of difficulties in AN: eating symptoms and social avoidance. Sixty adolescent girls with AN were recruited following hospitalisation in a specialised department. They completed self-administered questionnaires of alexithymia (TAS-20), of central symptoms of the eating disorders (EDI), and of anxious and depressive affects (SCL-90). Anxiety and social avoidance were assessed in the course of a standardised interview (LSAS). These measures were performed at inclusion, and at 6-, 12- and 18-months' follow-up. The relationship between TAS-20 and EDI or LSAS total scale scores across the four time points was assessed using mixed-effects models, including anxiety, depression, BMI, anorexia subtype, and age as co-factors. Partial least square regression was used to refine this multivariate analysis at subscale level, at inclusion and 18 months. Robust associations between TAS-20 and EDI scores were found, independently from anxious and depressive scores, nutritional state and AN subtype. These effects appeared more particularly linked to the implication of the dimensions difficulties identifying and describing feelings, interpersonal mistrust, feelings of inadequacy and interoceptive awareness deficit. There was also a durable association between alexithymia and social anxiety and avoidance, after adjusting for the confounding effects of depression, and anxiety, and the state of starvation. Difficulties in describing feelings appeared particularly involved here. Thus alexithymia does appear as a factor in the persistence of disorders in AN, and difficulties identifying and describing feelings could compound the social difficulties and major the relational isolation of these patients.

Anorexia nervosa (AN) involves serious risk for the patients concerned, and in severe cases the condition can be life-threatening. Indeed AN ranks first amongst life-threatening psychiatric pathologies [1]. Numerous factors are implicated in the onset of AN, amongst which socio-affective difficulties are particularly prominent in the most recent aetiological models [2]. As early as the 1960s, Hilde Bruch highlighted the difficulties experienced by anorexic patients in perceiving or controlling their bodily sensations, and confusion in the identification of their mental states [3] and [4]. These deficits in interoception and emotional introspection in AN have since been explored further via studies of alexithymia, a multifaceted personality construct that has emerged in line with the idea that deficits in the ability to experience and symbolise emotion can have adverse effects on well-being. The core alexithymia dimensions are: difficulties in identifying feelings and distinguishing between feelings and bodily sensations of emotional arousal; difficulties in describing feelings to other people; reduced daydreaming and constricted imaginal processes; a stimulus-bound, externally oriented, cognitive style [5]. As expected, studies in this area have indeed shown a marked prevalence of alexithymia amongst AN subjects, reporting rates of between 40% and 77% depending on the study, whilst the prevalence is approximately 10% amongst the controls of AN patients or in studies on healthy subjects [6], [7], [8], [9], [10], [11], [12], [13], [14], [15] and [16]. High prevalence of alexithymia is also reported in the two studies that specifically concerned anorexic adolescents [17] and [18]. However, it is important to note that in some of these studies it is difficult to disentangle the effect of the AN diagnosis from that of comorbid disorders, in particular anxiety and depression, which have also been linked to alexithymia [5], [19], [20], [21], [22] and [23]. This is also the case in many studies that used a dimensional approach [6], [10], [16], [18], [24], [25], [26], [27], [28] and [29]. Across 21 samples of anorexic patients published on this issue to our knowledge, only 8 studies to date have taken account of this possible confounding effect [11], [12], [30], [31], [32], [33], [34] and [35]. Four of these studies were able to establish that their AN patients were more alexithymic than the controls after adjusting for the level of depression [11], [30], [34] and [35], one of them showed that this was also true after adjusting for both the level of anxiety and depression [31], but three no longer found any difference between AN subjects and controls after adjustment on depression and anxiety scores [12], [32] and [33]. Hence, as for other psychiatric disorders with comorbid depression and/or anxiety, the question remains whether, in anorexia nervosa, alexithymia is a stable trait or a mental state secondary to depression or negative affectivity. Besides the level of negative affectivity, nutritional status is another issue that might have a critical impact on results. This can be exemplified by a recent study [34] which compared 14 restrictive AN (mean BMI: 17.9, SD: 1.8), 17 bingeing-purging AN (mean BMI: 17.8, SD: 2.0) and 22 controls (mean BMI: 21.8, SD: 2.5). In the overall sample, alexithymia scores (TAS-20) and BMI were significantly and negatively correlated (−.35). In addition, when BMI and depression were adjusted for in the group comparisons, the bingeing-purging AN remained significantly more alexithymic than the controls, but the difference between the restrictive AN and the controls disappeared. Another argument suggesting that the patients' nutritional status could account for certain apparently inconsistent results stems from a careful examination of the literature on AN and alexithymia. In fact, in the studies in favour of an independent role for alexithymia in AN (i.e. existing irrespective of dysphoric affects), the nutritional state of the participants was very severe (mean BMI ranging from 14.9 to 15.4 [11], [30] and [35]), whereas in studies where the relationship with alexithymia was no longer observed after adjustment on dysphoric affects, the mean BMI of participants was higher (i.e. respectively 16.6 and 17.2 [12] and [32]). Regarding the alexithymia construct, as it is multi-dimensional, comprising both affective and cognitive dimensions, another issue worth investigating is whether it is all its dimensions or only some that have a role in AN. Studies performed in student populations suggest that it might be more specifically the difficulties in identifying feelings (DIF) and in describing feelings (DDF) that are linked to an increased risk of developing eating disorders [36] and [37]. Amongst anorexic subjects, the DIF dimension could be associated with poor outcome according to one study [38]. In addition, three studies on clinical populations that explored the links between scores on an alexithymia measure (TAS-20, [39]) and eating symptoms (EDI, [40]) showed that the dimensions DIF and DDF were associated with the severity of the eating disorder. In these studies, associations have been reported with interoceptive awareness deficit, feelings of inadequacy and interpersonal mistrust, these being central symptom domains that are associated with serious forms of AN [7], [28] and [35]. However, some of these studies have serious limitations, not only because the possible effects of dysphoric affects were not taken into account (see [28] for a discussion on this issue), but above all on account of a major psychometric bias [7] and [35]. Indeed the instruments classically used to assess alexithymia (TAS and TAS-20) and eating symptoms (EDI) partially overlap. Certain items in the Toronto Alexithymia scales (TAS and TAS-20) are directly derived from the EDI interoceptive awareness scale. Thus an adjusted score needs to be calculated for accurate study of the links between alexithymia and eating symptoms using TAS scales and EDI questionnaire (see [28] for further details). Intriguingly, although alexithymia has been linked to both intrapersonal and interpersonal effects [41] and [42], the question of whether alexithymia is related to prosocial skills in AN is still poorly understood. It is common to observe difficulties in adapting socially in AN [43] and [44], which are described in particular in professional or school environments, or in leisure activities [43] and [45]. High levels of anxiety and social avoidance [46], and a strong prevalence of social phobias [47] and [48] have also been reported, which might contribute to impairing social adjustment [45]. We believe that alexithymia could be critically linked to these poor social adjustment features in AN. This hypothesis is supported by existing clinical studies amongst individuals with social phobias, which showed that they have difficulties identifying and describing their feelings (as measured on the TAS-20) [49] and [50]. Another argument is provided by a recent study amongst 319 students [51], which found positive correlations between alexithymia scores (TAS-20) and social anxiety or avoidance (as measured by a self-report version of the Liebowitz social anxiety scale [52]). Regarding the different dimensions of alexithymia, it is the difficulties in describing feelings in particular that were found to be predictive of anxiety and avoidance scores (on the basis of hierarchical linear regression adjusting on depression and anxiety [51]). However to our knowledge, no study to date has formally explored the links between alexithymia and anxiety/social avoidance amongst AN subjects. The main aim of this research was to take account of the methodological limitations of earlier studies so as to contribute to determining how far alexithymia might be linked to the severity of eating and interpersonal symptoms in AN. To do this we used the data from a large sample of anorexic adolescent girls followed over a period of 18 months after hospitalisation, taking account of the potentially confounding effect of dysphoric affects (depression and anxiety) and nutritional status, and also allowing for the metric overlap of the instruments used (TAS/EDI). The secondary objective was to determine the contribution of the different dimensions of alexithymia to the severity of eating and interpersonal symptoms in AN.