ترس و اجتناب از ارتباط چشمی در اختلال اضطراب اجتماعی

Abstract Background Excessive fear of scrutiny is a defining feature of social anxiety disorder. Eye contact may trigger feelings of being scrutinized, and although eye contact is commonly feared in persons with social anxiety disorder, it has been studied little. The purpose of this study was to characterize fear and avoidance of eye contact in patients with social anxiety disorder and in nonpatient samples. Methods Gaze fears and avoidance, social anxiety, and depression were assessed in 44 patients with generalized social anxiety disorder, 17 matched healthy comparison subjects, and 79 undergraduates. Patients were reassessed after 8 to 12 weeks of treatment with paroxetine. A new self-report instrument, the Gaze Anxiety Rating Scale (GARS), was used to assess fear and avoidance of eye contact, and its psychometric properties were analyzed. Results Patients with generalized social anxiety disorder, in comparison with healthy control participants, reported significantly increased levels of fear and avoidance of eye contact, which decreased significantly after 8 to 12 weeks of treatment with paroxetine. Fear and avoidance of eye contact were significantly associated with severity of social anxiety in all 3 samples. The GARS demonstrated excellent internal consistency within each sample. Conclusions Self-reported fear and avoidance of eye contact are associated with social anxiety in both nonpatient and social anxiety disorder samples. Preliminary psychometric analyses suggest that the GARS has utility in the assessment of gaze anxiety.

2. Results 2.1. Reliability of the GARS In the undergraduate sample, Cronbach α coefficients were .88 for the anxiety subscale, .90 for the avoidance subscale, and .95 for the full scale. Item-total correlations (using the full 34-item score with the particular item removed) ranged from 0.30 to 0.71, all significant (P < .05). Anxiety and avoidance scales were highly correlated (r = 0.75, P < .001). Among GSAD patients, Cronbach's α coefficients were .91 for the anxiety subscale, .93 for the avoidance subscale, and .96 for the full scale. Item-total correlations (using the full 34-item score with the particular item removed) ranged from 0.31 to 0.73. Anxiety and avoidance scales were highly correlated (r = 0.92, P < .001). Among the HC participants, coefficients were .85 for the anxiety subscale, .81 for the avoidance subscale, and .91 for the full scale. The test-retest reliability of the GARS was high in the subset of HC participants who were randomly selected to be reassessed after an 8-week interval (r = 0.99, P < .001), with Spearman correlations yielding identical results. 2.2. Gaze anxiety in student, GSAD, and HC samples In the undergraduate sample, mean (SD) total scores on the GARS did not differ between men and women (25.6 [14.8] vs 29.9 [16.2]; df = 76, t = 1.2, P = .25), although the range of total scores in men (4-49) did not extend as far into the severe end as did the range in women (4-70). Mean (SD) total scores on the GARS did not differ between whites and nonwhites (28.8 [16.6] vs 27.4 [13.8]; df = 76, t = 0.37, P = .72), and they were not associated with age (r = 0.08, P = .47). The GSAD patients reported significantly increased levels of gaze anxiety on the GARS compared with the HC participants, with large between-group effect sizes that approached the magnitude of that found for the global measure of social anxiety (LSAS, see Table 1). Among the 37 patients who completed open label paroxetine treatment, gaze anxiety decreased significantly (see Table 2), as did severity of social anxiety, submissive behavior, and depressive symptoms. Among the GSAD patients at baseline, total scores on the GARS did not differ between men and women (mean [SD], 45.8 [21.6] vs 54.6 [19.6]; df = 42.0, t = 1.4, P = .17), although the range in men (7-91) extended further into the milder end than did the range in women (26-89). Total scores in the HC sample ranged from 0 to 24, with only 1 subject reporting a total score of higher than 15. Mean (SD) total scores on the GARS for the GSAD patients did not differ between whites and nonwhites (49.1 [20.9] vs 51.3 [21.5]; df = 36.0, t = 0.34, P = .73), and scores were not associated with age (r = 0.20, P = .19), with similarly nonsignificant findings in the HC group. Table 1. Baseline assessments for HC and GSAD participants HC GSAD df t P d Mean SD Mean SD GARS 6.7 6.7 50.0 20.9 57.9 12.2 <.001 2.8 Anxiety 4.3 4.0 26.6 9.9 58.8 12.5 <.001 3.0 Avoidance 2.4 3.0 23.4 11.4 55.0 11.2 <.001 2.5 LSAS 7.9 5.3 77.1 20.2 53.6 20.7 <.001 4.7 SBS 13.6 6.8 35.9 9.0 59.0 9.2 <.001 2.8 HAM-D 0.3 0.7 7.1 4.7 47.5 9.4 <.001 2.0 BDI 1.0 1.5 13.3 11.3 46.8 7.1 <.001 1.5 HC indicates healthy comparison; GSAD, generalized social anxiety disorder; GARS, Gaze Anxiety Rating Scale; LSAS, Liebowitz Social Anxiety Scale; SBS, Submissive Behavior Scale; HAM-D, Hamilton Depression Scale; BDI, Beck Depression Inventory. Table options Table 2. Baseline and posttreatment assessments for GSAD completers of 8 to 12 weeks of paroxetine Baseline Posttreatment df t P d Mean SD Mean SD GARS 49.1 20.2 24.9 18.1 34 6.0 <.001 1.3 Anxiety 26.4 9.5 14.8 9.3 34 5.6 <.001 1.2 Avoidance 22.6 11.1 10.1 9.7 34 5.9 <.001 1.2 LSAS 77.6 18.9 41.4 25.3 34 8.2 <.001 1.6 SBS 35.0 9.0 26.2 8.7 37 6.3 <.001 1.0 HAM-D 6.9 4.8 3.6 3.6 37 3.4 .002 0.8 BDI 11.6 9.4 5.0 6.0 36 4.9 <.001 0.8 GSAD, generalized social anxiety disorder; GARS, Gaze Anxiety Rating Scale; LSAS, Liebowitz Social Anxiety Scale; SBS, Submissive Behavior Scale; HAM-D, Hamilton Depression Scale; BDI, Beck Depression Inventory. Table options On the additional descriptive items of the GARS, the GSAD patients, in comparison with the HC participants, reported increased rates of at least moderate severity of gaze avoidance due to anxiety (54.6% vs 0%; χ2 = 13.1, df = 59, P < .0001), feeling self-conscious making eye contact (61.3% vs 0%; χ2 = 16.3, P < .0001), and difficulty deciding how much eye contact is best (50% vs 0%; χ2 = 11.2, df = 59, P < .001). Both groups reported low rates of gaze avoidance attributed to interference with concentration (11.4% vs 0%; P = .31) and fear of staring too much into others' eyes (20.5% vs 0%; P = .27). Most GSAD and HC participants viewed making eye contact as at least moderately important for their social and work relationships (75% vs 82%; P = .74). The GSAD patients reported onset of fear or avoidance of eye contact at a mean (SD) age of 12.9 (5.1) years, not significantly different from their mean (SD) age of onset of SAD (14.6 [6.6] years; t = 1.3, df = 30, P = .20). 2.3. Correlations with social anxiety, depression, and submissive behavior In the undergraduate sample, GARS total scores were significantly correlated with social anxiety (SIAS, r = 0.65, P < .001). Depression and submissive behavior were not assessed in this sample. In the GSAD patients at baseline, GARS total scores were significantly correlated with measures of social anxiety (LSAS, r = 0.61, P < .001) and submissive behavior (SBS, r = 0.47, P = .001) but not with measures of depression (HRSD, r = 0.21, P = .17; BDI, r = 0.26, P = .09.). The GARS-LSAS correlation was significantly greater than correlations of the GARS with the HRSD (z = 2.5, P = .01) and with the BDI (z = 2.1, P = .03). Spearman correlations showed the same pattern of significance except that the GARS-LSAS correlation was greater than the GARS-BDI correlation at only a trend level of significance (z = 1.8, P = .07). In the HC sample at baseline, GARS total scores were significantly correlated with social anxiety and with the BDI (LSAS, r = 0.74, P = .001; SBS, r = 0.19, P = .47; HRSD, r = 0.03, P = .90; BDI, r = 0.56, P = .02). The only significant difference between correlations found the GARS-LSAS correlation to be greater than the GARS-HRSD correlation (z = 2.7, P = .01). Spearman correlations showed the same pattern of significance except that the correlation of the GARS with the BDI was not significant (rs = 0.27, P = .34).