Alexithymia may be considered as a personality feature characterized by poorness of imaginary life, speech focused on actual facts and physical sensations, general inaccuracy in or paucity of the words used to express emotions, and recourse to acting out to avoid intrapsychic conflicts. The possible link between alexithymia and psychosomatic or psychopathological disorders is now well documented. In particular, studies suggested that alexithymia may be frequently observed in obese or bulimic patients. This study was designed to investigate the link between obesity and alexithymia according to the presence or not of binge eating problems; 40 obese female patients (BMI > or = 27.3) seeking obesity treatment and 32 normal weight women used as controls were included in the study. In the obese group, 11 patients (27.5%) exhibited binge-eating disorder according to the DSM IV criteria. Alexithymia was assessed using the Toronto Alexithymia Scale (TAS), and past and current mental disorders were assessed by means of the Structured Clinical Interview for DSM III-R (SCID). In addition, current depression was assessed using the Beck Depression Inventory (BDI). The mean TAS score was found significantly higher in obese patients than in controls (72.6 +/- 11.8 vs 65.2 +/- 9.3, respectively; p < 0.005). In the same way, alexithymia (defined by TAS score > or = 74) was found significantly more frequent in obese patients than in controls (52.5% vs 21.8%, respectively; p < 0.03). However, among obese patients no significant difference was found between patients with and without binge-eating disorder. Current major depression was also found significantly more frequent in obese patients than in controls (15% vs 0%, respectively; p < 0.03), and the mean BDI score was very significantly higher in obese patients (12.2 +/- 8.7 vs 4.6 +/- 4.6, respectively; p < 0.0001). Comparisons between obese patients with and without binge-eating disorder showed that only past major depression was found significantly more frequent in those with binge-eating disorder (81.8% vs 10.3%, respectively; p < 0.0001), although the mean BDI score was significantly higher in patients with binge-eating disorder (18.5 +/- 11.7 vs 9.8 +/- 5.9, respectively; p < 0.02). Group by group comparisons suggested that two factors may play a role in the correlation found between obesity and alexithymia. First, the mean TAS score was found significantly higher in subjects with low educational level (p < 0.05), obese patients exhibiting significantly lower educational level when compared to controls (p < 0.002). Then, a significant positive correlation was found between TAS scores and BDI scores (Spearman's test: p < 0.01), obese patients showing significantly higher BDI scores than controls (p < 0.0001). In order to confirm these results, a logistic regression procedure was performed in the total sample (obese patients + controls). Three factors were found significantly increasing the risk to get a TAS score > or = 74: low educational level (odds ratio: 3.56), past and/or current major depression (odds ratio: 2.77), and BDI score > or = 8 (odds ratio: 2.18). Obesity in itself had no significant effect on TAS scores. Our results confirm that alexithymia is a psychological feature frequently observed in obese patients. In our study, the correlation found between obesity and alexithymia appears to be irrespective of binge-eating disorder, and seems to be mediated by the educational level and the frequency of associated depression. However, further investigations need to be done in order to specify the relationships between obesity, alexithymia, low educational level, and depression.