Borderline Personality Disorder
Conditions
Keywords
Dialectical behavior therapy, Emotions, Behavior and Behavior mechanisms, Randomized Controlled Trials
Brief summary
Although dialectical behavior therapy (DBT) skills training is effective in the treatment of borderline personality disorder, it contains four skills modules and there is little research to guide their modular application. This study compares the unique effects of two distinct DBT skills training modules, relative to a non-DBT therapy group for adults with borderline personality disorder. Using innovative laboratory-based assessment methods, the proposed study will examine the effects of these conditions on emotional responding and interpersonal functioning, as well as clinical outcomes.
Detailed description
Borderline personality disorder (BPD) is a severe mental health condition with high morbidity and mortality. Although dialectical behavior therapy (DBT) is an efficacious treatment for BPD, it is resource-intensive and lengthy in its full form, involving one year of weekly individual therapy and group skills training in mindfulness, emotion regulation, interpersonal effectiveness, and distress tolerance. As a result, few patients have access to the full treatment. A better understanding of how the distinct components of DBT affect different sets of symptoms could help to streamline this treatment and personalize its use with specific patients. Improvements in both interpersonal and emotional functioning are theorized to underlie improvements in BPD. Thus, emotion regulation and interpersonal effectiveness skills training may be particularly important components of DBT. Therefore, this study examines the unique effects of two distinct DBT skills training modules. Participants are adults with BPD and recent, recurrent self-injurious behaviors (planned N = 81) who are randomly assigned to six weeks of DBT emotion regulation skills training (DBT-ER), DBT interpersonal effectiveness skills training (DBT-IE), or a non-skills control group. Using innovative laboratory-based multimethod assessments, this study examines the effects of these conditions on emotional responding and interpersonal functioning, as well as BPD related outcomes. Aim 1 examines the unique effects of DBT-ER and DBT-IE on their respective emotion-related (subjective and biological emotional reactivity, behavioral emotion regulation, skills use) and interpersonal (subjective and behavioral) targets, compared to the non-DBT treatment. Aim 2 examines whether improved emotional functioning predicts reductions in BPD symptoms and self-injury. Aim 3 examines whether baseline emotion dysregulation interacts with treatment condition to predict treatment response. The proposed research is innovative in its experimental examination of the effects of DBT components on specific targets in BPD. Given the high societal costs of BPD, this work has important public health significance. Findings will inform larger studies evaluating the potential modular use of DBT components to result in briefer and more efficient individualized treatments for patients.
Interventions
BEHAVIORALDialectical behavior therapy - emotion regulation skills training
Arm 1. Dialectical behavior therapy - emotion regulation skills training follows the emotion regulation skills DBT Skills Training Manual Second Edition and the DBT Skills Training Handouts and Worksheets Second Edition. This group involves 6 weekly sessions. Arm 2. Dialectical behavior therapy - interpersonal effectiveness skills training follows the interpersonal effectiveness skills DBT Skills Training Manual Second Edition and the DBT Skills Training Handouts and Worksheets Second Edition. This group involves 6 weekly sessions.
BEHAVIORALDialectical behavior therapy - interpersonal effectiveness skills training
Arm 2. Dialectical behavior therapy - interpersonal effectiveness skills training follows the interpersonal effectiveness skills DBT Skills Training Manual Second Edition and the DBT Skills Training Handouts and Worksheets Second Edition. This group involves 6 weekly sessions.
BEHAVIORALInterpersonal psychotherapy
Arm 3. Non-skills-oriented interpersonal psychotherapy group follows evidence-based principles on common factors in a group therapy context. This group involves 6 weekly sessions.
Sponsors
National Institute of Mental Health (NIMH)
University of Toledo
University of Massachusetts, Amherst
Study design
Allocation
RANDOMIZED
Intervention model
PARALLEL
Primary purpose
BASIC_SCIENCE
Masking
NONE
Eligibility
Sex/Gender
ALL
Age
18 Years to 60 Years
Healthy volunteers
No
Inclusion criteria
1. exhibit 4+ BPD symptoms, 2. have a history of recent (i.e., past-year) and recurrent (\> 1 instance) of self-injury, 3. commit to participate in one of our 6-week experimental groups, 4. have an individual health provider who can manage imminent issues, 5. be between 18-60 years old,
Exclusion criteria
1. not fluent in English, 2. have impaired (uncorrected) vision or hearing that would impair ability to understand study stimuli, 3. a current manic, psychotic, or active physiological dependence on substances (to limit interference in the lab), 4. low cognitive functioning (IQ ≤ 70.4 (TOPF; Pearson Assessments, 2009), 5. past DBT treatment
Design outcomes
Primary
| Measure | Time frame | Description |
|---|---|---|
| Change in Emotional Functioning - Quad | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14 | Assessed with the Difficulties in Emotion Regulation Scale (DERS), which has total scores that range from 36-180, with higher scores indicating more difficulties. Outcomes are reported as slopes (scores over weeks, centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models. This measure is listed separately to permit presentation of two units of measure (that are in the same model) - the Linear unit of measure (DERS scores/week), and the Quadratic unit of measure (DERS scores/week\^2). This measure reflects the Quadratic measure. |
| Change in Borderline Personality Disorder Features - Quad | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14 | Assessed with the abbreviated Borderline Symptom List (BSL23), which has mean scores that range from 0-4, with higher scores indicating more symptoms. Outcomes are reported as slopes (scores rescaled \*100 over weeks, centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models. This measure is listed separately to permit presentation of two units of measure (that are in the same model) - the Linear unit of measure (BSL scores/week), and the Quadratic unit of measure (BSL scores/week\^2). This measure reflects the Quadratic measure. |
| Change in Emotional Functioning - Linear | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14 | Assessed with the Difficulties in Emotion Regulation Scale (DERS), which has total scores that range from 36-180, with higher scores indicating more difficulties. Outcomes are reported as slopes (scores over weeks, centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models. This measure is listed separately to permit presentation of two units of measure (that are in the same model) - the Linear unit of measure (DERS scores/week), and the Quadratic unit of measure (DERS scores/week\^2). This measure reflects the Linear measure. |
| Change in Borderline Personality Disorder Features - Linear | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14 | Assessed with the abbreviated Borderline Symptom List (BSL23), which has mean scores that range from 0-4, with higher scores indicating more symptoms. Outcomes are reported as slopes (scores rescaled \*100 over weeks, centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models. This measure is listed separately to permit presentation of two units of measure (that are in the same model) - the Linear unit of measure (BSL scores/week), and the Quadratic unit of measure (BSL scores/week\^2). This measure reflects the Linear measure. |
Secondary
| Measure | Time frame | Description |
|---|---|---|
| Change in Emotional Habituation | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7 | Assessed with skin conductance (SC; microsiemens, μS) in response to emotional cues presented in the lab (of note, per protocol, we examined DV in response to negative and neutral image stimuli \[Neg\] across initial \[Novel\] and repeated presentations) which were presented along with sound bursts to elicit startle. Data were transformed to address skew (μStr). Outcomes are reported as novel x negative slopes (SLP) over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects / multilevel models. |
| Change in Deliberate Physiological Emotional Regulation | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7 | Assessed with high frequency heart rate variability (HRV, ms2/hz) in response to emotional scripts (Negative: negative vs. neutral) with emotion regulation instructions (Reg: Maintain vs. Decrease in the Negative condition) presented in the lab. Data were transformed to address skew (ms2/hztr). Outcomes are reported as slopes of the effect of the Reg condition on HRV over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects /multilevel models. |
| Change in Physiological Emotional Reactivity | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7 | Assessed with skin conductance (SC; microsiemens, μS) in response to emotional scripts (Neg: neutral, negative) by emotion regulation (Reg: emotional maintain instructions, and negative with decrease instructions) presented in the lab. Data were transformed to address skew (μStr). Outcomes are reported as slopes of Negative (Neg) condition effects on SC over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects / multilevel models. |
| Change in Self-reported Emotional Reactivity | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7 | Assessed with self-reported emotions on the PANAS negative affect (NA) scale (mean scores on a scale of 1-5) in response to emotional scripts (Neg: neutral (Neu), negative (Neg)) by emotion regulation condition (Reg: emotional maintain instructions, and negative with decrease instructions) presented in the lab. Outcomes are reported as slopes differences in PANAS NA scores between neutral and negative-maintain over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models. |
| Change in Self-reported Emotional Regulation | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7 | Assessed with self-reported emotions on the PANAS negative affect (NA) scale (mean scores on a scale of 1-5) in response to emotional scripts (neutral (Neu), negative (Neg)) by emotion regulation condition (Reg: with emotional maintain (Main) instructions, and negative with decrease (Dec) instructions) presented in the lab. Outcomes are reported as slopes differences in PANAS NA scores between Reg conditions in response to negative stimuli (negative-decrease and negative-maintain) over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/ multilevel models. |
| Change in Affect-modulated Startle | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7 | Assessed with eyeblink startle amplitude (mV) in response to emotional cues presented in the lab (of note, per protocol, we examined eyeblink in response to negative and neutral image stimuli across initial \[novel\] and repeated presentations) which were presented along with sound bursts to elicit startle. This outcome was transformed and then rescaled (\*100, = mVtr). Outcomes are reported as slopes per condition over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models with image valence (Neg: negative vs. neutral) and type (Novel: novel vs. repeated) within-person condition effects. |
Other
| Measure | Time frame | Description |
|---|---|---|
| Change in Coping Strategies - Linear | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14 | Assessed with the DBT-Ways of Coping Checklist (DBT-WCCL, transformed = DBTWCCLtr), which yields scales of skills use (DSS, rescaled \*100), general dysfunctional coping (DC1, rescaled \*10), and blaming others (DC2, \*10), with scores of 0-3, higher reflects greater use. Outcomes are reported as slopes (with SE) centered around treatment start per condition, extracted from mixed effects/multilevel models (separately per scale score).This measure is listed separately to permit presentation of two units of measure (that are in the same model) for the DC1 scale, which was fit best with a quadratic model, whereas the remaining scales only required a Linear model. This measure reflects the Linear measures (DBTWCCL/week) for all scales. |
| Change in Coping Strategies - Quad | Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14 | Assessed with the DBT-Ways of Coping Checklist (DBT-WCCL, transformed = DBTWCCLtr), which yields scales of skills use (DSS, rescaled \*100), general dysfunctional coping (DC1, rescaled \*10), and blaming others (DC2, \*10), with scores of 0-3, higher reflects greater use. Outcomes are reported as slopes (SE) centered around treatment start per condition, extracted from mixed effects/multilevel models (separately per scale score).This measure is listed separately to permit presentation of two units of measure (that are in the same model) for the DC1 scale, which was fit best with a quadratic model, whereas the remaining scales only required a Linear model. This measure reflects the Quadratic measure (DBTWCCL/week\^2). |
Countries
United States
Participant flow
Recruitment details
N=115 recruited from the area around western MA assessed for eligibility but only 84 enrolled in study. .
Pre-assignment details
Due to group design, participants waited for next group date to begin treatment
Participants by arm
| Arm | Count |
|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training Dialectical behavior therapy - emotion regulation skills training
Dialectical behavior therapy - emotion regulation skills training: Arm 1. Dialectical behavior therapy - emotion regulation skills training follows the emotion regulation skills DBT Skills Training Manual Second Edition and the DBT Skills Training Handouts and Worksheets Second Edition. This group involves 6 weekly sessions.
Arm 2. Dialectical behavior therapy - interpersonal effectiveness skills training follows the interpersonal effectiveness skills DBT Skills Training Manual Second Edition and the DBT Skills Training Handouts and Worksheets Second Edition. This group involves 6 weekly sessions. | 27 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training Dialectical behavior therapy - interpersonal effectiveness skills training
Dialectical behavior therapy - interpersonal effectiveness skills training: Arm 2. Dialectical behavior therapy - interpersonal effectiveness skills training follows the interpersonal effectiveness skills DBT Skills Training Manual Second Edition and the DBT Skills Training Handouts and Worksheets Second Edition. This group involves 6 weekly sessions. | 28 |
| Non-skills-oriented Interpersonal Psychotherapy Group Non-skills-oriented interpersonal psychotherapy group
Interpersonal psychotherapy: Arm 3. Non-skills-oriented interpersonal psychotherapy group follows evidence-based principles on common factors in a group therapy context. This group involves 6 weekly sessions. | 29 |
| Total | 84 |
Withdrawals & dropouts
| Period | Reason | FG000 | FG001 | FG002 |
|---|---|---|---|---|
| Overall Study | 4-miss from treatment | 6 | 5 | 5 |
| Overall Study | Withdrawal by Subject | 1 | 0 | 3 |
Baseline characteristics
| Characteristic | Total | Dialectical Behavior Therapy - Emotion Regulation Skills Training | Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Non-skills-oriented Interpersonal Psychotherapy Group |
|---|---|---|---|---|
| Age, Categorical <=18 years | 0 Participants | 0 Participants | 0 Participants | 0 Participants |
| Age, Categorical >=65 years | 0 Participants | 0 Participants | 0 Participants | 0 Participants |
| Age, Categorical Between 18 and 65 years | 84 Participants | 27 Participants | 28 Participants | 29 Participants |
| Age, Continuous | 25.61 years STANDARD_DEVIATION 9.65 | 26.41 years STANDARD_DEVIATION 11.71 | 25.75 years STANDARD_DEVIATION 9.95 | 24.72 years STANDARD_DEVIATION 7.21 |
| DERS total | 118.688 units on a scale STANDARD_DEVIATION 21.363 | 118.525 units on a scale STANDARD_DEVIATION 23.707 | 116.539 units on a scale STANDARD_DEVIATION 21.399 | 120.915 units on a scale STANDARD_DEVIATION 23.707 |
| Ethnicity (NIH/OMB) Hispanic or Latino | 8 Participants | 5 Participants | 2 Participants | 1 Participants |
| Ethnicity (NIH/OMB) Not Hispanic or Latino | 76 Participants | 22 Participants | 26 Participants | 28 Participants |
| Ethnicity (NIH/OMB) Unknown or Not Reported | 0 Participants | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) American Indian or Alaska Native | 0 Participants | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Asian | 13 Participants | 3 Participants | 5 Participants | 5 Participants |
| Race (NIH/OMB) Black or African American | 2 Participants | 2 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) More than one race | 6 Participants | 4 Participants | 2 Participants | 0 Participants |
| Race (NIH/OMB) Native Hawaiian or Other Pacific Islander | 0 Participants | 0 Participants | 0 Participants | 0 Participants |
| Race (NIH/OMB) Unknown or Not Reported | 6 Participants | 3 Participants | 1 Participants | 2 Participants |
| Race (NIH/OMB) White | 57 Participants | 15 Participants | 20 Participants | 22 Participants |
| Region of Enrollment United States | 84 participants | 27 participants | 28 participants | 29 participants |
| Sex: Female, Male Female | 75 Participants | 24 Participants | 26 Participants | 25 Participants |
| Sex: Female, Male Male | 9 Participants | 3 Participants | 2 Participants | 4 Participants |
Adverse events
| Event type | EG000 affected / at risk | EG001 affected / at risk | EG002 affected / at risk |
|---|---|---|---|
| deaths Total, all-cause mortality | 0 / 27 | 0 / 28 | 0 / 29 |
| other Total, other adverse events | 0 / 27 | 0 / 28 | 0 / 29 |
| serious Total, serious adverse events | 0 / 27 | 0 / 28 | 0 / 29 |
Outcome results
Primary
Change in Borderline Personality Disorder Features - Linear
Assessed with the abbreviated Borderline Symptom List (BSL23), which has mean scores that range from 0-4, with higher scores indicating more symptoms. Outcomes are reported as slopes (scores rescaled \*100 over weeks, centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models. This measure is listed separately to permit presentation of two units of measure (that are in the same model) - the Linear unit of measure (BSL scores/week), and the Quadratic unit of measure (BSL scores/week\^2). This measure reflects the Linear measure.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14
Population: missing 1 data point for DBT-ER group
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Borderline Personality Disorder Features - Linear | -1.716 BSL scores/wk | Standard Error 0.808 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Borderline Personality Disorder Features - Linear | -3.567 BSL scores/wk | Standard Error 0.746 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Borderline Personality Disorder Features - Linear | -3.667 BSL scores/wk | Standard Error 0.833 |
Comparison: Comparisons of Linear Slopes (DBT-ER vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.09Mixed Models Analysis
Comparison: Comparisons of Linear Slopes (DBT-ER vs. DBT-IE) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.93Mixed Models Analysis
Comparison: Comparisons of Linear Slopes (DBT-ER vs. DBT-IE) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.93Mixed Models Analysis
Comparison: Comparisons of Quadratic Slopes (DBT-ER vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.93Mixed Models Analysis
Comparison: Comparisons of Quadratic Slopes (DBT-IE vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.836Mixed Models Analysis
Comparison: Comparisons of Quadratic Slopes (DBT-IE vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.994Mixed Models Analysis
Primary
Change in Borderline Personality Disorder Features - Quad
Assessed with the abbreviated Borderline Symptom List (BSL23), which has mean scores that range from 0-4, with higher scores indicating more symptoms. Outcomes are reported as slopes (scores rescaled \*100 over weeks, centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models. This measure is listed separately to permit presentation of two units of measure (that are in the same model) - the Linear unit of measure (BSL scores/week), and the Quadratic unit of measure (BSL scores/week\^2). This measure reflects the Quadratic measure.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14
Population: missing 1 data point for DBT-ER group from ITT sample
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Borderline Personality Disorder Features - Quad | -.162 BSL scores/wk^2 | Standard Error 0.108 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Borderline Personality Disorder Features - Quad | -.163 BSL scores/wk^2 | Standard Error 0.095 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Borderline Personality Disorder Features - Quad | -.135 BSL scores/wk^2 | Standard Error 0.1 |
Comparison: Comparisons of Quadratic Slopes (DBT-ER vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.93Mixed Models Analysis
Comparison: Comparisons of Quadratic Slopes (DBT-IE vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.836Mixed Models Analysis
Comparison: Comparisons of Quadratic Slopes (DBT-IE vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.994Mixed Models Analysis
Primary
Change in Emotional Functioning - Linear
Assessed with the Difficulties in Emotion Regulation Scale (DERS), which has total scores that range from 36-180, with higher scores indicating more difficulties. Outcomes are reported as slopes (scores over weeks, centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models. This measure is listed separately to permit presentation of two units of measure (that are in the same model) - the Linear unit of measure (DERS scores/week), and the Quadratic unit of measure (DERS scores/week\^2). This measure reflects the Linear measure.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14
Population: ITT sample (see study description)
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Emotional Functioning - Linear | -.917 DERS scores/wk | Standard Error 0.31 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Emotional Functioning - Linear | -1.25 DERS scores/wk | Standard Error 0.32 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Emotional Functioning - Linear | -1.01 DERS scores/wk | Standard Error 0.26 |
Comparison: Comparisons of Linear Slopes (DBT-ER vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.8Mixed Models Analysis
Comparison: Comparisons of Linear Slopes (DBT-IE vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.544Mixed Models Analysis
Comparison: Comparisons of Linear Slopes (DBT-ER vs. DBT-IE) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.442Mixed Models Analysis
Primary
Change in Emotional Functioning - Quad
Assessed with the Difficulties in Emotion Regulation Scale (DERS), which has total scores that range from 36-180, with higher scores indicating more difficulties. Outcomes are reported as slopes (scores over weeks, centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models. This measure is listed separately to permit presentation of two units of measure (that are in the same model) - the Linear unit of measure (DERS scores/week), and the Quadratic unit of measure (DERS scores/week\^2). This measure reflects the Quadratic measure.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14
Population: ITT sample (see study description)
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Emotional Functioning - Quad | -.06 DERS scores/wk^2 | Standard Error 0.04 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Emotional Functioning - Quad | -.04 DERS scores/wk^2 | Standard Error 0.04 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Emotional Functioning - Quad | -.07 DERS scores/wk^2 | Standard Error 0.03 |
Comparison: Comparisons of Quadratic Slopes (DBT-ER vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.762Mixed Models Analysis
Comparison: Comparisons of Quadratic Slopes (DBT-IE vs. IP) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.478Mixed Models Analysis
Comparison: Comparisons of Quadratic Slopes (DBT-ER vs. DBT-IE) in latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) with quadratic effects.p-value: 0.705Mixed Models Analysis
Secondary
Change in Affect-modulated Startle
Assessed with eyeblink startle amplitude (mV) in response to emotional cues presented in the lab (of note, per protocol, we examined eyeblink in response to negative and neutral image stimuli across initial \[novel\] and repeated presentations) which were presented along with sound bursts to elicit startle. This outcome was transformed and then rescaled (\*100, = mVtr). Outcomes are reported as slopes per condition over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models with image valence (Neg: negative vs. neutral) and type (Novel: novel vs. repeated) within-person condition effects.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7
Population: Some participants were missing data due to nonattendance at lab sessions
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Affect-modulated Startle | .015 Condition diff (NegXNovel) in Mn mVtr/Wk | Standard Error 0.041 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Affect-modulated Startle | .043 Condition diff (NegXNovel) in Mn mVtr/Wk | Standard Error 0.04 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Affect-modulated Startle | -.076 Condition diff (NegXNovel) in Mn mVtr/Wk | Standard Error 0.044 |
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. The Negative x Novel x time (weeks, centered around the start of treatment) slopes were calculated per condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to IP.p-value: 0.13Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. The Negative x Novel x time (weeks, centered around the start of treatment) slopes were calculated per condition. Models tested whether these slopes differed by condition. This test compared DBT-IE to IP.p-value: 0.046Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. The Negative x Novel x time (weeks, centered around the start of treatment) slopes were calculated per condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to DBT-IE.p-value: 0.617Mixed Models Analysis
Secondary
Change in Deliberate Physiological Emotional Regulation
Assessed with high frequency heart rate variability (HRV, ms2/hz) in response to emotional scripts (Negative: negative vs. neutral) with emotion regulation instructions (Reg: Maintain vs. Decrease in the Negative condition) presented in the lab. Data were transformed to address skew (ms2/hztr). Outcomes are reported as slopes of the effect of the Reg condition on HRV over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects /multilevel models.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Deliberate Physiological Emotional Regulation | .08 Condition diff (Reg) on Mn ms2/hztr/week | Standard Error 0.1 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Deliberate Physiological Emotional Regulation | .10 Condition diff (Reg) on Mn ms2/hztr/week | Standard Error 0.1 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Deliberate Physiological Emotional Regulation | .10 Condition diff (Reg) on Mn ms2/hztr/week | Standard Error 0.09 |
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of the effect of negative maintain vs. negative decrease on weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to IP.p-value: 0.211Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of the effect of negative maintain vs. negative decrease on weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-IE to IP.p-value: 0.148Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of the effect of negative maintain vs. negative decrease on weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to DBT-IE.p-value: 0.857Mixed Models Analysis
Secondary
Change in Emotional Habituation
Assessed with skin conductance (SC; microsiemens, μS) in response to emotional cues presented in the lab (of note, per protocol, we examined DV in response to negative and neutral image stimuli \[Neg\] across initial \[Novel\] and repeated presentations) which were presented along with sound bursts to elicit startle. Data were transformed to address skew (μStr). Outcomes are reported as novel x negative slopes (SLP) over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects / multilevel models.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7
Population: Missing data due to nonattendance at lab visits
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Emotional Habituation | -.156 Condition diff (NegxNovel) in Mn μStr/wk | Standard Error 0.07 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Emotional Habituation | .023 Condition diff (NegxNovel) in Mn μStr/wk | Standard Error 0.069 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Emotional Habituation | .134 Condition diff (NegxNovel) in Mn μStr/wk | Standard Error 0.076 |
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of change in this interaction effect were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to IP.p-value: 0.005Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of change in this interaction effect were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-IE to IP.p-value: 0.281Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of change in this interaction effect were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to DBT-IE.p-value: 0.066Mixed Models Analysis
Secondary
Change in Physiological Emotional Reactivity
Assessed with skin conductance (SC; microsiemens, μS) in response to emotional scripts (Neg: neutral, negative) by emotion regulation (Reg: emotional maintain instructions, and negative with decrease instructions) presented in the lab. Data were transformed to address skew (μStr). Outcomes are reported as slopes of Negative (Neg) condition effects on SC over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects / multilevel models.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Physiological Emotional Reactivity | 1.46 Condition diff (Neg) in Mean μS/week | Standard Error 0.48 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Physiological Emotional Reactivity | -.71 Condition diff (Neg) in Mean μS/week | Standard Error 0.4 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Physiological Emotional Reactivity | .41 Condition diff (Neg) in Mean μS/week | Standard Error 0.39 |
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of the effect of negative maintain vs. neutral images on weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to IP.p-value: 0.108Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of the effect of negative maintain vs. neutral images on weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-IE to IP.p-value: 0.014Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of the effect of negative maintain vs. neutral images on weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to DBT-IE.p-value: 0.001Mixed Models Analysis
Secondary
Change in Self-reported Emotional Reactivity
Assessed with self-reported emotions on the PANAS negative affect (NA) scale (mean scores on a scale of 1-5) in response to emotional scripts (Neg: neutral (Neu), negative (Neg)) by emotion regulation condition (Reg: emotional maintain instructions, and negative with decrease instructions) presented in the lab. Outcomes are reported as slopes differences in PANAS NA scores between neutral and negative-maintain over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/multilevel models.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Self-reported Emotional Reactivity | -.07 Condition diff (Neg-Neu) in NA scores/wk | Standard Error 0.06 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Self-reported Emotional Reactivity | -.09 Condition diff (Neg-Neu) in NA scores/wk | Standard Error 0.06 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Self-reported Emotional Reactivity | -.12 Condition diff (Neg-Neu) in NA scores/wk | Standard Error 0.06 |
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. The slopes of the difference between neutral and negative-maintain were regressed on time (weeks, centered around the start of treatment) and calculated per condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to IP.p-value: 0.475Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. The slopes of the difference between neutral and negative-maintain were regressed on time (weeks, centered around the start of treatment) and calculated per condition. Models tested whether these slopes differed by condition. This test compared DBT-IE to IP.p-value: 0.654Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. The slopes of the difference between neutral and negative-maintain were regressed on time (weeks, centered around the start of treatment) and calculated per condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to DBT-IE.p-value: 0.791Mixed Models Analysis
Secondary
Change in Self-reported Emotional Regulation
Assessed with self-reported emotions on the PANAS negative affect (NA) scale (mean scores on a scale of 1-5) in response to emotional scripts (neutral (Neu), negative (Neg)) by emotion regulation condition (Reg: with emotional maintain (Main) instructions, and negative with decrease (Dec) instructions) presented in the lab. Outcomes are reported as slopes differences in PANAS NA scores between Reg conditions in response to negative stimuli (negative-decrease and negative-maintain) over weeks (centered around the start of treatment, with standard errors) per condition, extracted from mixed effects/ multilevel models.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Self-reported Emotional Regulation | -.18 Condition diff (Reg) in NA score/wk | Standard Error 0.07 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Self-reported Emotional Regulation | -.01 Condition diff (Reg) in NA score/wk | Standard Error 0.07 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Self-reported Emotional Regulation | .15 Condition diff (Reg) in NA score/wk | Standard Error 0.07 |
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. We examined the effect of Negative Decrease vs. Negative Maintain over time per condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to IP.p-value: 0.001Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. We examined the effect of Negative Decrease vs. Negative Maintain over time per condition. Models tested whether these slopes differed by condition. This test compared DBT-IE to IP.p-value: 0.078Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. We examined the effect of Negative Decrease vs. Negative Maintain over time per condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to DBT-IE.p-value: 0.08Mixed Models Analysis
Other Pre-specified
Change in Coping Strategies - Linear
Assessed with the DBT-Ways of Coping Checklist (DBT-WCCL, transformed = DBTWCCLtr), which yields scales of skills use (DSS, rescaled \*100), general dysfunctional coping (DC1, rescaled \*10), and blaming others (DC2, \*10), with scores of 0-3, higher reflects greater use. Outcomes are reported as slopes (with SE) centered around treatment start per condition, extracted from mixed effects/multilevel models (separately per scale score).This measure is listed separately to permit presentation of two units of measure (that are in the same model) for the DC1 scale, which was fit best with a quadratic model, whereas the remaining scales only required a Linear model. This measure reflects the Linear measures (DBTWCCL/week) for all scales.
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14
| Arm | Measure | Group | Value (MEAN) | Dispersion |
|---|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Coping Strategies - Linear | DC1 - Linear - rescaled *10 | -.17 DBTWCCLtr scores/week | Standard Error 0.08 |
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Coping Strategies - Linear | DSS - Linear ONLY - rescaled *100 | .23 DBTWCCLtr scores/week | Standard Error 0.559 |
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Coping Strategies - Linear | DC2 - Linear ONLY- rescaled *10 | -.174 DBTWCCLtr scores/week | Standard Error 0.079 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Coping Strategies - Linear | DC1 - Linear - rescaled *10 | -.31 DBTWCCLtr scores/week | Standard Error 0.07 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Coping Strategies - Linear | DSS - Linear ONLY - rescaled *100 | .546 DBTWCCLtr scores/week | Standard Error 0.523 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Coping Strategies - Linear | DC2 - Linear ONLY- rescaled *10 | -.230 DBTWCCLtr scores/week | Standard Error 0.074 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Coping Strategies - Linear | DSS - Linear ONLY - rescaled *100 | .544 DBTWCCLtr scores/week | Standard Error 0.562 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Coping Strategies - Linear | DC2 - Linear ONLY- rescaled *10 | -.219 DBTWCCLtr scores/week | Standard Error 0.08 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Coping Strategies - Linear | DC1 - Linear - rescaled *10 | -.26 DBTWCCLtr scores/week | Standard Error 0.07 |
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DSS over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to IP.p-value: 0.693Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DSS over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-IE to IP.p-value: 0.998Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DSS over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to DBT-IE.p-value: 0.68Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DC1 over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-ER to IP.p-value: 0.333Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DC1 over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-IE to IP.p-value: 0.632Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DC1 over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared DBT-IE to DBT-ER.p-value: 0.178Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DC2 over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared the slope differences DBT-ER to IP.p-value: 0.689Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DC2 over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared the slope differences DBT-IE to IP.p-value: 0.917Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DC2 over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared the slope differences DBT-ER to DBT-IE.p-value: 0.603Mixed Models Analysis
Other Pre-specified
Change in Coping Strategies - Quad
Assessed with the DBT-Ways of Coping Checklist (DBT-WCCL, transformed = DBTWCCLtr), which yields scales of skills use (DSS, rescaled \*100), general dysfunctional coping (DC1, rescaled \*10), and blaming others (DC2, \*10), with scores of 0-3, higher reflects greater use. Outcomes are reported as slopes (SE) centered around treatment start per condition, extracted from mixed effects/multilevel models (separately per scale score).This measure is listed separately to permit presentation of two units of measure (that are in the same model) for the DC1 scale, which was fit best with a quadratic model, whereas the remaining scales only required a Linear model. This measure reflects the Quadratic measure (DBTWCCL/week\^2).
Time frame: Pre-treatment, mid-treatment week 3-4, post-treatment week 6-7, follow-up week 13-14
Population: ITT sample
| Arm | Measure | Value (MEAN) | Dispersion |
|---|---|---|---|
| Dialectical Behavior Therapy - Emotion Regulation Skills Training | Change in Coping Strategies - Quad | -002 DBTWCCLtr DC1 scores/week^2 | Standard Error 0.011 |
| Dialectical Behavior Therapy - Interpersonal Effectiveness Skills Training | Change in Coping Strategies - Quad | -.012 DBTWCCLtr DC1 scores/week^2 | Standard Error 0.007 |
| Non-skills-oriented Interpersonal Psychotherapy Group | Change in Coping Strategies - Quad | -.015 DBTWCCLtr DC1 scores/week^2 | Standard Error 0.007 |
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DC1 over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared the quadratic effects differed DBT-ER to IP.p-value: 0.313Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DC1 over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared the quadratic effects differed DBT-IE to IP.p-value: 0.747Mixed Models Analysis
Comparison: Latent growth curve models (in multilevel/mixed effects framework, Mplus v8.1) were estimated with maximum likelihood. Simple slopes of DC1 over weeks were calculated for each condition. Models tested whether these slopes differed by condition. This test compared the quadratic effects differed DBT-ER to DBT-IE.p-value: 0.444Mixed Models Analysis