تأثیر رفتاردرمانی دیالکتیکی در مقابل درمان جامعه توسط کارشناسان بر روی تجربه احساسی، بیان و پذیرش در اختلال شخصیت مرزی

Evidence suggests that heightened negative affectivity is a prominent feature of Borderline Personality Disorder (BPD) that often leads to maladaptive behaviors. Nevertheless, there is little research examining treatment effects on the experience and expression of specific negative emotions. Dialectical Behavior Therapy (DBT) is an effective treatment for BPD, hypothesized to reduce negative affectivity (Linehan, 1993a). The present study analyzes secondary data from a randomized controlled trial with the aim to assess the unique effectiveness of DBT when compared to Community Treatment by Experts (CTBE) in changing the experience, expression, and acceptance of negative emotions. Suicidal and/or self-injuring women with BPD (n = 101) were randomly assigned to DBT or CTBE for one year of treatment and one year of follow-up. Several indices of emotional experience and expression were assessed. Results indicate that DBT decreased experiential avoidance and expressed anger significantly more than CTBE. No differences between DBT and CTBE were found in improving guilt, shame, anxiety, or anger suppression, trait, and control. These results suggest that DBT has unique effects on improving the expression of anger and experiential avoidance, whereas changes in the experience of specific negative emotions may be accounted for by general factors associated with expert therapy. Implications of the findings are discussed.

.

Results Preliminary analyses Two models (random intercept, random intercept and random slope) and several covariance structures were tested for fit for each outcome variable and the best fitting model is reported in Table 2. Two potential confounding factors were examined for each model: treatment drop out and cohort effects. A pattern mixture model was employed where the main effect of each confound, confound by time, and confound by time by condition were added to each HLM model. If the main effect or any of the interaction effects was significant we kept these terms in the analyses and excluded the non-significant predictors. Completing the intervention was assessed for all outcomes; cohort effects were assessed for the STAXI and AAQ, the only measures administered to both cohorts of participants. Table 1. Descriptive statistics based on observed values at each assessment period for each outcome measure. Measure Stat. Pre treatment 4-Months 8-Months 12-Months 16-Months 20-Months 24-Months AAQ DBT M 80.12 75.18 73.86 67.20 67.62 64.98 64.53 SD 10.03 10.97 9.46 12.80 12.68 11.67 13.23 N 33 38 36 41 42 40 45 CTBE M 78.94 79.60 76.62 74.52 68.09 71.13 69.85 SD 12.50 11.33 12.41 12.68 12.98 14.88 13.08 N 32 28 26 31 32 32 34 STAXI Out DBT M 2.19 2.04 2.02 1.84 1.95 1.89 1.90 SD 0.75 0.46 0.66 0.51 0.49 0.53 0.43 N 52 49 49 48 46 44 45 CTBE M 1.99 1.95 1.80 1.93 1.82 1.99 1.89 SD 0.59 0.57 0.56 0.57 0.42 0.61 0.52 N 49 42 36 36 36 34 35 STAXI In DBT M 2.75 2.58 2.56 2.33 2.40 2.24 2.24 SD 0.60 0.58 0.48 0.67 0.65 0.63 0.57 N 52 49 49 48 46 44 45 CTBE M 2.76 2.59 2.33 2.40 2.40 2.36 2.34 SD 0.53 0.48 0.53 0.52 0.58 0.55 0.55 N 49 43 37 36 36 34 35 STAXI Control DBT M 2.53 2.59 2.63 2.78 2.83 2.78 2.75 SD 0.80 0.65 0.67 0.63 0.61 0.66 0.71 N 52 49 49 48 46 44 45 CTBE M 2.59 2.68 2.74 2.80 2.88 2.70 2.67 SD 0.68 0.78 0.64 0.59 0.65 0.65 0.61 N 49 42 36 36 36 34 35 STAXI Trait DBT M 2.29 2.23 2.06 1.98 1.97 1.93 1.97 SD 0.75 0.55 0.62 0.59 0.52 0.56 0.54 N 52 49 49 48 46 44 45 CTBE M 2.24 2.09 1.93 2.05 1.94 2.11 1.95 SD 0.66 0.70 0.58 0.56 0.49 0.60 0.55 N 49 43 37 37 36 34 35 PFQ Shame DBT M 2.19 2.20 2.00 1.90 1.91 1.72 1.87 SD 0.68 0.64 0.55 0.48 0.54 0.42 0.56 N 31 30 29 28 27 25 26 CTBE M 2.43 2.12 1.97 2.11 2.13 2.02 1.87 SD 0.64 0.50 0.74 0.63 0.62 0.55 0.66 N 31 26 23 23 22 20 21 PFQ Guilt DBT M 2.33 2.20 2.21 1.87 1.97 1.75 1.91 SD 1.00 0.95 0.72 0.81 0.84 0.74 0.73 N 31 30 29 28 27 25 26 CTBE M 2.29 2.15 1.85 1.99 1.92 1.93 1.77 SD 0.78 0.74 0.84 0.97 0.78 0.85 0.95 N 31 26 23 23 22 20 21 TMAS DBT M 0.73 0.69 0.66 0.61 0.58 0.54 0.55 SD 0.15 0.19 0.18 0.23 0.24 0.25 0.24 N 32 30 30 28 25 24 25 CTBE M 0.82 0.72 0.64 0.74 0.66 0.69 0.67 SD 0.09 0.15 0.22 0.15 0.21 0.19 0.19 N 31 26 23 22 22 18 20 Note. AAQ = Acceptance and Action Questionnaire, STAXI = State Trait Anger Inventory, PFQ = Personal Feelings Questionnaire, TMAS = Tailor Manifest Anxiety Inventory, DBT = Dialectical Behavior Therapy, CTBE = Community Treatment by Experts. Table options Table 2. HLM results for each outcome measure. Measure HLM slope for Slope estimate S.E. t a d.f. p Covariance structuresb AAQ DBT Cohort 1 −1.95 0.42 −4.65 92.95 <.001 Unstructured, Autoregressive DBT Cohort 2 −3.61 0.42 −8.70 67.14 <.001 CTBE Cohort 1 −0.86 0.45 −1.93 90.74 .06 CTBE Cohort 2 −2.51 0.45 −5.56 71.93 <.001 DBT vs. CTBE −1.09 0.50 −2.19 79.98 .03 Cohort 1 vs. 2 1.65 0.50 3.31 79.86 <.001 STAXI Anger Control DBT 0.04 0.01 3.24 86.13 <.005 Unstructured CTBE 0.02 0.01 1.22 92.11 .23 DBT vs. CTBE 0.02 0.02 1.26 89.48 .21 STAXI DBT −0.09 0.01 −6.86 79.51 <.001 Unstructured, Autoregressive Anger In CTBE −0.06 0.01 −4.22 83.00 <.001 DBT vs. CTBE −0.03 0.02 −1.46 81.50 .15 STAXI Anger Out DBT −0.05 0.01 −4.26 88.14 <.001 Unstructured CTBE −0.01 0.01 −0.90 96.69 .37 DBT vs. CTBE −0.04 0.02 −2.21 92.80 .03 STAXI Anger Trait DBT Drops −0.09 0.02 −4.35 98.62 <.001 Unstructured, Autoregressive DBT Completers −0.05 0.01 −3.94 81.20 <.001 CTBE Drops −0.08 0.02 −3.96 106.8 <.001 CTBE Completers −0.03 0.01 −2.30 80.48 .02 DBT vs. CTBE −0.02 0.02 −0.84 83.78 .40 Completers vs. Drops 0.05 0.02 2.16 99.74 .03 PFQ Shame DBT −0.07 0.02 −3.16 54.29 <.005 Unstructured, Autoregressive CTBE −0.08 0.02 −3.60 56.64 <.001 DBT vs. CTBE 0.01 0.03 0.48 55.63 .64 PFQ Guilt DBT −0.07 0.03 −2.89 51.27 <.01 Unstructured, Autoregressive CTBE −0.07 0.03 −2.63 52.62 .01 DBT vs. CTBE −0.001 0.04 −0.03 52.08 .98 TMAS DBT −0.03 0.01 −3.91 47.89 <.001 Unstructured CTBE −0.02 0.01 −2.79 47.71 .01 DBT vs. CTBE −0.01 0.01 −0.52 47.80 .60 Note. a Each slope is compared to a null slope using a t-test, and the t value, degrees of freedom and significance are presented in this table. b A random intercept and random slope HLM model with an unstructured covariance structure for the random effects provided the best fit for the data for all outcomes. To test whether a model accounting for within subject correlation between time-points better explains the data, we fit an autoregressive (AR1) correlated residual error structure to each model (Gibbons et al., 1993). This more complex model was selected only if the model fit was significantly improved. S.E. = standard error; AAQ = Acceptance and Action Questionnaire, STAXI = State Trait Anger Inventory, PFQ = Personal Feelings Questionnaire, TMAS = Tailor Manifest Anxiety Inventory, DBT = Dialectical Behavioral Therapy, CTBE = Community Treatment by Experts, Drops = participants who dropped out of treatment, Completers = participants who completed treatment. Table options There were no significant main effects of treatment drop out status, and there were no significant drop out by time, drop out by condition by time interaction effects for Experiential Avoidance, Anger In, Anger Out, Anger Control, Shame, or Guilt (p > .05), suggesting that drop out status did not play a significant role in these outcomes. For Anger Trait, there was a significant drop out (yes/no) by time interaction, which was included in the main Anger Trait analyses, F(1, 98.89) = 4.40, p < .05; there was no significant main effect of drop out, or significant time by condition by drop out interaction effect, F(1, 98.65) = 0.16, p = .69; F(1, 90.28) = 0.51, p = .48, respectively. There were no significant main effects of cohort, and no significant cohort by time or cohort by condition by time interaction effects for any of the Anger, Shame, or Guilt variables (p > .05). For Experiential Avoidance, there was a significant cohort by time interaction, which was included in the main analyses, F(1,78.79) = 11.83, p < .001. The main effect of cohort and the cohort by time by condition interaction effect were not significant predictors of Experiential Avoidance, F(1,77.20) = 0.24, p = .63; F(1,77.79) = 1.58, p = .21, respectively. Primary analyses Descriptive data and sample sizes for all outcome measures at each assessment point are provided in Table 1. Table 2 summarizes the HLM results related to model and slope coefficient. Anger. Analyses with the STAXI subscales (Anger In, Anger Out, Anger Control, Anger Trait) indicated that, across both treatment conditions, a significant effect of time in treatment was found, F (1, 81.50) = 59.70, p < .001; F (1, 92.80) = 12.46, p < .001; F (1, 89.48) = 9.44, p < .005; F (1, 101.33) = 35.91, p < .001, respectively. These results suggest that, overall, participants significantly improved over time on the four anger variables. A significant interaction between time and condition was found for Anger Out with DBT showing a significant reduction in anger expression whereas CTBE did not significantly change, F(1, 92.80) = 4.87, p < .05, d = 0.31, observed power = 0.30. Although no significant time by condition interaction was present for Anger Control, F(1, 89.48) = 1.58, p = .21, participants in DBT improved significantly, but participants in CTBE did not improve significantly in anger control (d = 0.19, observed power = 0.15, see Table 2). There were no significant interaction effects for Anger Trait (d = 0.13; observed power = 0.10) and Anger In (d = 0.31; observed power = 0.29). Participants who dropped out of treatment improved significantly more on Anger Trait scores than participants who completed treatment ( Table 2). Experiential avoidance. Analyses confirmed both hypotheses related to changes during treatment in experiential avoidance. A significant effect of time and a significant time by condition interaction effect were found, F (1, 79.98) = 78.90, p < .001; F (1, 79.98) = 4.78, p < .05, respectively. The difference in slope between DBT and CTBE indicates that experiential avoidance decreased at a faster rate in DBT than CTBE ( Table 2, d = 0.59, observed power was 0.57). The second cohort of participants had a faster reduction of experiential avoidance than the first cohort, with the CTBE participants in cohort two not improving significantly over time. Shame. For shame, a significant effect of time was found, F (1, 55.63) = 22.95, p < .001, but no significant condition by time interaction, F (1, 55.63) = 0.23, p = .64, d = 0.13, observed power = 0.07. The slope estimates for both DBT and CTBE conditions were significantly different from zero, indicating that both treatments resulted in significant improvements in shame ( Table 2). Guilt. For guilt, a significant effect of time, F (1, 52.08) = 15.17, p < .001, but no significant condition by time interaction was found, F (1, 52.08) = 0.001, p = .98, d = 0.01, observed power = 0.05. The slope estimates for both DBT and CTBE were significantly different from zero, which suggests that both treatments yielded significant improvements in guilt ( Table 2). Anxiety. Analyses confirmed a significant effect of time for anxiety, F (1, 47.80) = 21.89, p < .001, but no significant interaction, F (1, 47.80) = 0.27, p = .60, d = 0.27, observed power = 0.09. The slope estimates for both DBT and CTBE conditions were significantly different from zero, indicating that both treatments resulted in significant improvements in anxiety ( Table 2).