تفاوتهای جنسی در اثرات غیر مستقیم فرایندهای شناختی بر کاهش اضطراب از مشکلات تنظیم احساسات

Abstract Drawing on a gender roles theory of emotion regulation, we examined the specific facets of emotion regulation difficulties through which higher-order cognitive abilities may be related to anxiety. Participants (N = 225) completed self-report measures of emotion regulation difficulties and anxiety, and were administered neuropsychological tests assessing abstract reasoning and inhibition. PROCESS (Hayes, 2012) was used to estimate the direct and indirect effects of both inhibition and abstract reasoning on anxiety symptoms, with six dimensions of emotion regulation difficulties serving as multiple mediators operating in parallel. Results suggest that the relation between higher-order cognitive abilities and anxiety operate through distinct, sex-dependent emotion dysregulation mechanisms. For females, higher levels of inhibition and abstract thinking were associated with poorer clarity of emotions, which in turn, was associated with higher levels of anxiety symptoms. As such, over-attentiveness to, or over-analysis of, emotions may be particularly detrimental among females who have relatively higher abstract reasoning abilities. For males, higher inhibition was associated with greater perceived effectiveness in regulating negative emotions, which in turn, was associated with lower levels of anxiety symptoms. This finding suggests that mood regulation expectancies may be particularly important in understanding the pathogenesis of anxiety in males.

3. Results 3.1. Preliminary analyses and descriptive statistics Attrition analysis was conducted for all variables that were administered at T1. Results revealed no significant differences between T1 completers and those who completed both T1 and T2 assessment sessions on descriptive statistics (i.e., sex, race, year in school) and anxiety symptoms. In regard to emotion regulation difficulties, participants who completed both T1 and T2 had significantly higher scores on the DERS–Nonacceptance subscale than those who only completed T1, t(214) = −2.54, p = .012 (M = 2.01, SD = 0.93, and M = 1.66, SD = 0.61, respectively); no significant differences were observed for any other DERS subscale. Missing data (i.e., 8.5% of all data points) were imputed using expectation–maximization (EM) imputation. Data screening was conducted to examine univariate and multivariate normality. Univariate outliers (i.e., >3.3 SDs from the mean) were recoded to the next highest or lowest value ( Tabachnick & Fidell, 2007). Three cases that were identified as having undue influence on the primary analytic models (i.e., multivariate outliers defined as Mahalanobis distances >3.3 SDs from the mean; Tabachnick & Fidell, 2007) were removed from the sample. We explored whether any demographic variables (i.e., race, year in school) were significantly associated with our dependent variables (thus requiring inclusion in primary analyses as a covariate[s]; Tabachnick & Fidell, 2007). Because mediation models were examined separately for males and females based on study hypotheses, sex was not included in this preliminary analysis. Among potential covariates, race (dichotomized as White vs. all others) was significantly associated with variables of interest. White participants, compared to all other participants, had significantly higher scores on both DERS–Nonacceptance, t(211) = 2.83, p = .005 (M = 2.09, SD = 0.92, and M = 1.76, SD = 0.77, respectively), and DERS–Awareness, t(211) = 3.22, p = .001 (M = 2.78, SD = 0.86, and M = 2.40, SD = 0.86, respectively). Accordingly, race was included as covariate in subsequent analyses. Means and standard deviations were examined for variables of primary interest. Based on the cut scores for the BAI, and providing evidence for the presence of clinically-relevant levels of anxiety pathology, approximately 23% of the sample endorsed mild anxiety symptoms, 9% endorsed moderate anxiety symptoms, and 2% endorsed severe anxiety symptoms. As expected, females reported significantly higher levels of anxiety than males, t(211) = −3.01, p = .003 (M = 9.80, SD = 8.30, and M = 6.71, SD = 6.21, respectively). Significant differences were not observed in emotion regulation difficulties or in inhibition. However, females obtained significantly lower abstract reasoning scaled scores compared to males, t(211) = 2.63, p = .009 (M = 9.87, SD = 2.07, and M = 10.59, SD = 1.89, respectively). 3.2. Mediation analyses PROCESS (Hayes, 2012) was used to estimate the direct and indirect effects of both inhibition and abstract reasoning on anxiety symptoms, with the individual DERS subscales serving as multiple mediators operating in parallel. Mediation models were examined separately for males and females and abstract reasoning and inhibition each served as the independent variable in separate models, thus resulting in a total of four parallel multiple mediator models. In all models, all DERS subscales were entered simultaneously to allow for the investigation of the indirect effects of each DERS subscale while controlling for the effect of the other DERS subscales. Race and color-naming speed were included as covariates in these analyses, as well as the alternative independent variable (e.g., when examining the direct and indirect effects of inhibition on anxiety, abstract reasoning was controlled for and vice versa). First, we examined the effects of both inhibition and abstract reasoning on anxiety symptoms through DERS subscales in females (see Fig. 1). The direct effects from both inhibition and abstract reasoning to anxiety were non-significant (β = .09 and β = −.06, respectively). For females, the only indirect effect to emerge for the relation between inhibition and anxiety symptoms was through DERS–Clarity (β = .07, 95% CI [.01, .22]). This indirect effect was such that higher levels of inhibition were associated with poorer clarity of emotions, which in turn, was associated with higher levels of anxiety symptoms. For abstract reasoning in females, a similar pattern emerged; the only indirect effect that emerged was through DERS–Clarity (β = .11, 95% CI [.02, .26]). Consistent with the pattern described above, higher levels of abstract thinking were associated with poorer clarity of emotions, which in turn, was associated with higher levels of anxiety symptoms. Path models of the relation between (a) inhibition and (b) abstract reasoning ... Fig. 1. Path models of the relation between (a) inhibition and (b) abstract reasoning and anxiety symptoms through emotion regulation difficulties in females. Race, color-naming speed, and the alternative cognitive variable (i.e., inhibition or abstract reasoning) are included as covariates. Path coefficients are unstandardized (standardized coefficients in parentheses). ∗p < .05. ∗∗p < .01. Figure options In the second set of meditation analyses, which was conducted for males (see Fig. 2), the direct effects from both inhibition and abstract reasoning to anxiety were non-significant (β = −.11 and β = .05, respectively). A significant indirect effect emerged for the relation between inhibition and anxiety symptoms through DERS–Strategies (β = −.13, 95% CI [−.31, −.03]), such that higher levels of inhibition were associated with higher levels of perceived access to effective emotion regulation strategies, which in turn, was associated with lower levels of anxiety symptoms. None of the other DERS subscales mediated the relationship between inhibition and anxiety. Additionally, none of the DERS subscales mediated the relationship between abstract thinking and anxiety symptoms in males. Path models of the relation between (a) inhibition and (b) abstract reasoning ... Fig. 2. Path models of the relation between (a) inhibition and (b) abstract reasoning and anxiety symptoms through emotion regulation difficulties in males. Race, color-naming speed, and the alternative cognitive variable (i.e., inhibition or abstract reasoning) are included as covariates. Path coefficients are unstandardized (standardized coefficients in parentheses). ∗p < .05. ∗∗p < .01. Figure options To test whether the aforementioned indirect effects significantly differed between males and females (i.e., sex as a moderator of the indirect effects), point estimates for each specific indirect effect (ab) were compared. Specifically, the difference between the ab paths was divided by the standard error of the estimate to form a z statistic (see Fairchild & MacKinnon, 2009). As noted by Fairchild and MacKinnon (2009), “If the point estimates in each group are statistically different from one another, there is significant moderation of the indirect effect (i.e., heterogeneity in the ab product), such that the mediated effect is moderated by group membership” (pp. 7, 8). The indirect effect for the relation between abstract reasoning and anxiety through DERS–Clarity was significantly moderated by sex (z = 1.78, p = .037). For the relation between inhibition and anxiety symptoms, the indirect effect through DERS–Clarity was not significantly moderated by sex (z = 1.27, p = .101), though the indirect effect through DERS–Strategies was significantly moderated by sex (z = −1.72, p = .043). Thus, evidence of moderated meditation emerged for the indirect pathways from abstract reasoning to anxiety through DERS–Clarity, as well as from inhibition to anxiety through DERS–Strategies.