Skip to main content
NCBI home page
NIHPA Author Manuscripts logoLink to NIHPA Author Manuscripts
. Author manuscript; available in PMC: 2019 Aug 1.
Published in final edited form as: Prof Psychol Res Pr. 2018 Aug;49(4):274–281. doi: 10.1037/pro0000200

Parental Validation and Invalidation Predict Adolescent Self-Harm

Molly Adrian 1,2, Michele S Berk 3, Kathryn Korslund 4, Kathryn Whitlock 2, Elizabeth McCauley 1,2,4, Marsha Linehan 4
PMCID: PMC6424515  NIHMSID: NIHMS984972  PMID: 30906109

Abstract

This study was designed to evaluate family processes theoretically implicated in the onset and maintenance of adolescent self-harm. In the present study, we focus on understanding parental validation and invalidation in response to their adolescent in order to estimate the association between parental responses and self-harm in a high risk group of adolescents. We also sought to determine the influence of psychotherapy on parental validation and invalidation over time during participation in a randomized clinical trial of psychotherapy designed to reduce self-harm. Thirty-eight teens (Mage= 14.85; 94.1% female, 55.3% Caucasian, and 17.5% Latino) and their parents participated in three assessments over a six month period corresponding to pretreatment, midtreatment and end of treatment in the trial. Results indicate a robust association between parental validation, invalidation and adolescent self-harm. There were no significant associations observed between parental validation, invalidation, and adolescent suicidal ideation. Observed levels of parental validation and invalidation were not changed during the six-month course of psychotherapy.

Keywords: Self-harm, parental validation, parent-child interactions, adolescent psychotherapy

Suicide is a leading cause of death among individuals between the ages of 10 and 34 in the United States (Curtin, Warner & Hedegaard, 2016). Rates of suicide increase exponentially from childhood into adolescence and have increased over the past 20 years (Curtin, Warner & Hedegaard, 2016). Prior self-harm, including suicide attempts (SA) and non-suicidal self-injury (NSSI), is a reliable predictor for future self-harm (Asarnow et al., 2011; Klonsky, May, & Glenn, 2013). Despite this substantial impact of self-harm behaviors on families, healthcare systems, and communities, research to date has not established proven strategies to lower the rates of suicide (Curtin et al., 2016). Given these data, self-harm is a formidable threat to adolescent wellbeing. Attention to the family processes that are theoretically linked to the development and maintenance of adolescent self-harm is an important research priority.

Family Processes are Associated with Self-Harm

Contextual family processes studied in relation to self-harm have included family history of suicide, abuse and neglect, family dysfunction/conflict, and poor parent-child communication (for a review, see Bridge, Goldstein & Brent, 2006). A prominent theory in understanding the development of self-harm is the biosocial theory (Linehan, 1993). According to the biosocial theory (Linehan, 1993), severe and persistent difficulty regulating emotions is seen as the primary contributing factor related to self-harm. Difficulty regulating emotion is thought to develop in childhood based on the transaction between a biological predisposition to emotional reactivity on the part of the child, and an invalidating interpersonal environment. The invalidating environment is defined as one in which communication of emotion is met with caregiver responses that are inconsistent, inappropriate to the emotion expressed, and/or minimize the significance of the emotional experience. As a result, the child, who is already prone to strong emotional reactions, does not develop adequate emotion regulation skills as opportunities for co-regulation and teaching emotional de-escalation are missed. In this model, the function of self-harm is to manage severe emotion dysregulation in the absence of more constructive coping strategies (e.g., Adrian et al., 2011). Within this paradigm, adolescents who experience invalidating responses from the environment may also respond by escalating emotions and behaviors in an attempt to be taken more seriously, creating a coercive cycle of interactions that fuel emotional reactivity and self-harm (Crowell, Beauchaine, & Linehan, 2009). If the association between invalidation and self-harm was empirically supported, then reduction of parental invalidation would be a logical target for change in parent-adolescent interactions to reduce self-harm.

Data Supporting Biosocial Theory of Self-Harm

A small body of experimental and developmental research supports the core components of the biosocial theory. Not surprisingly, poor quality family relationships have been investigated broadly in relation to self-harm (Garber, Little, Hilsman, & Weaver, 1998; Johnson et al., 2002). There is growing evidence suggesting that negative family processes may play a role in self-harm. For example, Wedig and Nock (2007) found that high levels of parental expressed emotion, including criticism and emotional over-involvement, were associated with the presence and frequency of adolescent self-harm. Moreover, the authors found that parental criticism interacted with self-criticism to predict the highest rates of self-harm (Wedig & Nock, 2007). Crowell and colleagues (2008, 2013) conducted a series of studies that utilized an observational coding system to understand parent-teen interactions among teens with and without a history of self-harm. The authors reported that families of self-injuring adolescents demonstrated less positive affect, more negative affect, and lower cohesiveness compared to non-injuring controls (Crowell et al., 2008). Shenk and Fruzzetti (2011) randomly assigned groups of college age women to receive validating versus invalidating responses, and found distinct trajectories of emotional reactivity across both subjective and physiological measures consistent with the notion that invalidation impacts emotional reactivity (Linehan, 1993). On average, participants in the invalidating condition had significantly higher levels of negative affect, heart rate, and skin conductance when compared to individuals in the validating condition. The effect of invalidating responses on emotional reactivity was evident in analyses comparing both between-group differences in emotional reactivity as well as within-group change relative to baseline (Shenk & Fruzzetti, 2011). This pattern indicates that, on average, invalidating responses increased emotional reactivity during a stressful situation. Increased emotional reactivity resulting from invalidating responses may make regulation of such reactivity more difficult, especially for those individuals with limited emotion regulation skills, and potentiate the use of problematic behaviors to reduce such reactivity (Brain, Haines, & Williams, 1998; Nock & Mendes, 2008).

The Present Study

Based on the biosocial theory (Linehan, 1993), increasing validation and decreasing invalidation are core therapeutic tasks for treatment of adolescents at risk for self-harm. The purpose of the present study is to measure parental validation and invalidation, and the relationship between these constructs and self-harm behaviors, in a sample of youth at high risk for suicide. Subjects in this study were participants in a larger randomized controlled trial (RCT) examining the efficacy of Dialectical Behavior Therapy (DBT), as compared to Individual and Group Supportive Therapy (IGST), for decreasing self-harm (McCauley et al., in press). To the best of our knowledge, there are no prior studies of parental validation during adolescent treatment for self-harm. Both treatments used in the RCT, DBT and I/GST, addressed the adolescent’s experience of validation but had different principles for the target and methods for intervention. DBT is grounded in balancing acceptance and change strategies as a core principle in the treatment with explicit coaching in parental validation strategies (Miller et al., 2006); whereas I/GST focuses solely on clinician’s validation of the teen experience as the source of therapeutic action (Cohen et al., 2006). As reported with the full sample in McCauley and colleagues (in press) and Linehan and colleagues (2016), both conditions improved in treatment with respect to self-harm, with a significant advantage for DBT through the end of treatment (6 months), suggesting small but significant effects on self-harm outcomes.

In the present study, we examined levels of parental validation and invalidation using an observational coding task developed in prior research on DBT (Fruzzetti, 2010) in order to maximize ecological validity (Adrian & Berk, 2018). Parents and teens participated in a family conflict discussion task at three time points corresponding to pre-treatment, mid-treatment, and end of treatment during the larger RCT in order to measure validating and invalidating parental behaviors over time. There were four primary aims related to the parental behaviors and self-harm. First, we estimated baseline levels of parental validation and invalidation in this sample of youth at high risk for suicide. Second, we evaluated changes in validation and invalidation over time. Third, we estimated the association of baseline observed parental validation and invalidation with self-harm at the end of treatment (combining NSSI and SA in order to increase power). We hypothesized that lower levels of parental validation and higher levels of invalidation would be associated with greater self-harm. We also sought to evaluate if the association between validation and invalidation was significant for suicidal ideation. Because the biosocial theory specifically articulates parental invalidation affecting self-harm and not suicidal ideation, we hypothesized that the association between parental invalidation and suicidal ideation would not substantially impact treatment outcomes and thus not have significant associations between baseline parental behavior and end of treatment suicidal ideation severity. Finally, we conducted a preliminary exploration of the trajectory of parental validation and invalidation over the course of treatment. We hypothesized that parents of adolescents randomized to DBT would show greater increases in use of validation and decreases in the use of invalidation when discussing emotionally-evocative events as compared to parents of adolescents randomized to the I/GST condition due to explicit teaching of validation skills to parents in DBT.

Method

Participants

Participants included 38 adolescents aged 12–18 who were eligible for study procedures due to enrollment in a multisite randomized controlled trial evaluating the efficacy of DBT and IGST in reducing suicidal and self-harm behaviors (Collaborative Adolescent Research on Emotions and Suicide; see Berk et al., 2014). Inclusion criteria for the larger RCT included current suicidal ideation, repetitive self-harm, a history of suicide attempt, at least 2 borderline personality disorder features; exclusion criteria included psychotic disorders, autism spectrum diagnosis, and intellectual disability (McCauley et al., in press). The subsample, recruited from a single study site, was predominantly females (94.1%) and mean age was 14.85 years (SD=1.50). Race/ethnicity of participants was identified as follows: 2.6% Asian, 39.5% biracial, 2.6% other race/ethnicity, 55.3% white. Seventeen percent identified as Hispanic/Latino. Thirteen percent were born outside of the United States. Of those approached, 75% agreed to study procedures. Participants received $25 as a thank you for participating at each assessment.

Materials

Issues checklist.

The modified revealed differences task (Strodtbeck, 1951) is a 44-item scale that probes the frequency and intensity of the disagreements between parent and adolescent on a range of issues. The designated parent and adolescent completed this measure separately and rated the frequency, rated from 1 (never) to 5 (very often), and intensity, rated between 0 (calm) and 40 (very intense), for each issue listed.

Self-harm.

The primary outcome of interest was self-harm. The Suicide Attempt Self-Injury Count followed by the Suicide Attempt Self-Injury Interview (Linehan et al., 2006) measured the topography, intent, medical severity, social context, and outcomes of self-harm. Three trained assessors blind to study hypotheses and treatment condition collected information from the adolescent regarding the spectrum of self-harm behaviors for the six-month period from the first treatment session to the end of treatment (~6 months). In order to increase power, we combined SA and NSSI together into a single self-harm variable. This approach has been used in numerous prior studies on suicidal and self-harm behavior in adolescents (Hawton et al., 2012, Melhum et al., 2014), due to the high overlap between these behaviors, high association of both with subsequent suicide attempts and low base rate of suicide attempts in small to medium samples of suicidal youth.

Suicidal ideation.

The Suicide Ideation Questionnaire-Junior (Reynolds, 1988) was used to measure past-month suicide ideation. Although designed for younger adolescents, the SIQ-Jr has been widely used with both younger and older adolescents to evaluation suicide ideation. Use of the SIQ-Jr for all participants allows for measurement to be constant and minimize both participant burden (15 items) and reading comprehension problems. A raw score of 31 or above has been empirically established as the clinical cutoff indicating “potentially significant suicide risk,” with higher scores indicating more persistent suicidal ideation.

Procedures

All study procedures were approved by the University of Washington’s Institutional Review Board. Adolescents and their parents reported on stress and suicidality before and after the study procedures. Assessors were trained to intervene to reduce suicidality as needed. Following screening assessment to determine eligibility, adolescents and their parent completed a comprehensive assessment battery. Youth completed a questionnaire assessing suicidal ideation and both adolescents and parents completed structured interviews assessing adolescent self-harm. Teens and parents were next asked to engage in an interaction task. These procedures were conducted at baseline and repeated at intervals corresponding to assessments at mid treatment (3 months), and treatment completion (6 months). The measures used in this sub-study are described below. Following baseline assessment, youth were randomized to Individual/Group Supportive Therapy (I/GST) or Dialectical Behavior Therapy (DBT). In this sample, 52.6% (N=20) were assigned to IGST and 47.4% (N=18) were assigned to DBT.

Interaction task.

The items that were rated as moderately contentious based on the modified revealed differences task for the dyad were selected and presented to the dyad as the topic for the interaction task. Following the selection of the topic for discussion, parent and teen were asked to discuss the topic as they normally would, with a focus on how to solve the issue for approximately 10-minutes. The data from the interaction task was coded for parent validation and invalidation. Validation and invalidation are coded independently as individuals can demonstrate both validating and invalidating behaviors. Trained research assistants completed the coding using the Validating & Invalidating Behavior Coding Scale Manual (Fruzzetti, 2010) by Dr. Fruzzetti as the gold standard for training. The Validating and Invalidating Behavior Coding Scale is a valid and reliable measure of relationship functioning, and has been used in prior research (Fruzzetti, 2010; Shenk & Fruzzetti, 2011). Coders made ratings of observable behavior moment by moment to yield a global score for both validation and invalidation. There were seven levels of overall validation and invalidation for each interaction task, with higher score reflecting higher levels of the observed behavior. Participants were given global scores of validating and invalidating behaviors, interclass correlations ranged from .72 for validation to .88 for invalidation, consistent with the rating observed in the training manual (Fruzzetti, 2010).

Treatment conditions.

Dialectical Behavior Therapy (DBT) is the most extensively studied approach to reduction of both self-harm in adults (Clarkin, Levy, Lenzenweger, & Kernberg, 2007; Linehan et al., 2002; Linehan, Heard, & Armstrong, 1993; McMain et al., 2009; van den Bosch, Verheul, Schippers, & van den Brink, 2002). Results for adolescents have been slower to emerge but have shown promise for DBT as an effective intervention (Linehan et al., 2016; Mehlum et al., 2016; Miller, Rathus, & Linehan, 2006). As noted previously, the conceptual model of self-harm in DBT is based on a biosocial model (Linehan, 1993) whereby biological emotion vulnerability transacts with environmental invalidation to yield disordered functioning of both the adolescent and the environment. Adolescents in this condition received standard DBT including individual therapy, multifamily skills group, telephone coaching, and therapists participated in a weekly consultation team meeting. In DBT with adolescents, increasing parental validation and decreasing parental invalidation of teens are key treatment targets (Miller, Rathus & Linehan, 2007).

Individual and Group Supportive Therapy (I/GST) is an adaptation of the supportive therapy treatment manual developed by Brent and his colleagues (Brent et al., 1997), extended by Judy Cohen (TF CBT Trial), and modified by adding a group component to it match the provision of group therapy in DBT. Additionally, the treatment included use of protocols to explicitly target safety and stabilization of suicidal adolescents (See Berk et al., 2014 for more information). The conceptual model of self-harm utilized in I/GST is that adolescent self-harm is driven by the experience of feeling isolated, misunderstood, unloved and unwanted and hence; that a positive relationship with a therapist who provides unconditional positive regard will provide a corrective experience that reduces the need for self-harm. I/GST also included a group component in which teens engaged in supportive interactions with other teens while completing pleasant activities, in order to increase feelings of belongingness and acceptance. I/GST utilized a non-directive approach to support the adolescent’s self-esteem and belief in his/her inherent abilities to solve his/her own problems. Reduction of self-harm was conceptualized to come by therapeutic interactions that allowed the adolescent to trust and validate themselves. Up to seven parent-only sessions were allowed, in order to validate and increase their sense of capability related to parenting. Because the parent sessions did not explicitly teach or practice the ways in which parents could validate their adolescent’s experience, we did not expect parental responses towards their child to change.

Statistical Methods

Baseline parental validation and invalidation scores and their interaction were entered into (a) a Poisson model predicting self-harm (suicide attempts and non-suicidal self-injury) at the end of treatment (6 month assessment) and (b) a linear regression model predicting suicidal ideation at the end of treatment. When not significant, the interaction term was removed and model refit to determine statistical significance of validation and invalidation main effects. Missing baseline parental validation scores (N=8) were replaced with non-missing 3 month parental validation scores. Sensitivity analyses were conducted to examine the potential impact of high outlying numbers of self-harm in which (a) reports of >10 acts were replaced with 10; (b) participants reporting >10 acts were excluded. Similarly, in secondary analyses, the linear regression model predicting suicidal ideation was repeated with missing suicidal ideation values imputed using the Markov Chain Monte Carlo method with 10 imputations. This approach has been shown to be superior to listwise deletion (Sterne et al., 2009). Consistent with recommendations by Cohen, Cohen, West, and Aiken (2003),continuous predictors were mean centered prior to forming the interaction term. Significant interactions were probed with simple slopes analysis to examine the strength of the relationships between independent variable (i.e., validation) and the dependent variables (i.e., self-harm) at each level of the moderator (invalidation; Dawson, 2014).

To examine the predictive role of time and treatment on parental validation and invalidation, global ratings of validation and invalidation at baseline, midtreatment, and end of treatment assessments were examined with linear mixed modeling in R (R. Core Team, 2008) using the lme4 package (Bates, Maechler, & Bolker, 2013; Raudenbush & Bryk, 2002). Mixed modeling is flexible in its treatment of time as a continuous factor, allowing for variability in the actual time of assessment for each participant. In addition, mixed modeling can accommodate incomplete data across time making analyses more powerful due to the inclusion of more data points. All observations at the three timepoints of parental validation and invalidation were entered. Each equation included time, treatment, and their interaction. Time measured in days from the start of treatment was modeled using linear and quadratic terms, allowing for nonlinear change. Time will be modeled as a level 1variable. Treatment was dummy-variable, coded 0 for DBT and 1 for I/GST, and modeled as a level 2 variable. The cross-level interactions between Treatment and Time will model treatment differences in the trajectories across time. The Level 2 variance term models heterogeneity in individual participant trajectories. This analysis allowed for exploration of the changes in overall dimension ratings for validation and invalidation across treatment and time.

Results

Descriptive Statistics

For the 38 teen-parent dyads entering the study, 31 (81.5%) provided follow-up data. Two participants dropped out of the study, 2 were not able to be located, and 3 chose not to complete all assessment procedures. Descriptive data for suicidal ideation and self-harm are reported in Table 1. The data reveal almost ½ of the sample had a self-harm during the treatment period. Six percent of study participants reported 1 or more suicide attempt and 48.4% reported 1 or more non-suicidal self-injury acts (range 0 to 89) during the six month treatment period. All individuals who reported suicide attempts also engaged in NSSI. The mean suicidal ideation score was 36.5 (SD= 26.4) at the end of treatment (6 month assessment). Mean ratings of observed validation were level 3, showing attentive, functionally responding and clarifying validating behaviors. Mean ratings of observed invalidation were rated at level 3, showing missed opportunities and insisting as most commonly observed invalidating behaviors. There were no instances of the highest levels of validation, which matches self-disclosure in the context of another’s vulnerability, nor were there level 7 invalidating behavior which shows indifference to vulnerability (i.e., not responding at all to adolescent vulnerability).

Table 1.

Suicidal Ideation and Self-Harm as Reported by Adolescents at Baseline and 6 months

Item Baseline
N
Reporting
(%)		 M SD


range		 6 month
N
Reporting
(%)		 M SD range
Suicide Attempts 38 (100) 3.80 10.22 1–55 2 (6.4) 0.12 0.56 0–3
NSSI 38 (100) 34.26 44.30 1–173 15 (48.4) 7.19 19.33 0–89
Suicidal Ideation -- 44.67 24.9 24–90 -- 36.46 26.42 0–90
Observed Validation -- 3.08 1.57 1–6 -- 2.80 1.40 1–6
Observed
Invalidation		 -- 3.03 1.67 1–6 2.90 1.85 1–6
Open in a new tab

Effects of Observed Parental Behavior on Self-Harm

Associations of baseline parental validation and invalidation with self-harm and suicidal ideation at the end of treatment (6 months) are presented in Table 2. The association of higher levels of baseline parental validation with lower reported numbers of self-harm was moderated by higher levels of parental invalidation (p<0.0001). Follow-up analyses probed the two-way interaction between validation and invalidation. At high levels of validation, high parental invalidation was associated with dramatically higher numbers of self-harm (mean difference 21.0, 95% CI 6.8 to 64.9; p<.0001). At low levels of validation, the effect of invalidation on self-harm was attenuated (Figure 1). Suicidal ideation was not significantly associated with baseline parental validation (p=0.76) or invalidation (p=0.42). Results were similar in sensitivity analyses conducted to assess potential influence of outliers and missing values (details available upon request).

Table 2.

Regression Estimates for Validation and Invalidation Predicting Self-Injury Outcomes

Self-Harm
 Suicidal Ideation
β (SE) P Value β (SE) P Value
Intercept 12.49 (1.19) <.0001 35.49 (32.69) 0.28
Baseline Validation −4.93 (0.55) <.0001 −4.23 (7.28) 0.56
Baseline Invalidation −2.50 (0.30) <.0001 1.06 (8.09) 0.90
Validation * Invalidation 1.28 (0.14) <.0001 1.13 (2.31) 0.62
Open in a new tab

Note. Outcomes measured at 6 months. Results remained unchanged when sensitivity analyses were conducted (data available upon request).

Figure 1.

Figure 1.

Open in a new tab

The Interaction between Baseline Parental Validation and Invalidation and the Frequency of Self-Harm at 6-months

Effects of Treatment Assignment of Validation and Invalidation over Time

Hierarchical mixed effect models were used to estimate the predictive role of time and treatment on validation and invalidation at baseline, mid-treatment (3 months) and end of treatment (~6 months). Results of the full model, with predictors are displayed in Table 2. Neither time, treatment, nor the interaction of time and treatment predicted change in observed levels of parental validation or invalidation (β=.002, p=.49).

Discussion

Pinpointing predictors of adolescent self-harm is needed to help identify effective methods of suicide prevention for high-risk youth. The results of the present analyses indicate strong associations between observed parent validation and invalidation and subsequent self-harm behaviors. High levels of baseline parental validation interacted with low levels of invalidation to predict infrequent self-harm; however, at high levels of invalidation, self-harm was at its highest frequency. This same set of associations was not revealed with suicidal ideation, suggesting that parental invalidation may be uniquely predictive of self-harm. Our results augment Crowell and colleagues’ (2013) study in which microanalytic coding of parent-teen interactions identified that mothers of self-harming adolescents were more likely to invalidate, respond aversively, and matched, or escalated, conflict. In fact, mothers of self-harming teens only deescalated conflict in response to adolescent’s extreme behavior. Although Crowell and colleagues did not evaluate the association between levels of validation and invalidation with self-harm, their comparative study highlighted that the transactions between teens who harm themselves and their parents were significantly different than dyads without the history of self-harm. Additionally, in a longitudinal observational study utilizing a community sample of Chinese adolescents, adolescent-perceived family invalidation was significantly associated with the occurrence, but not frequency, of self-harm (You, Leung, Lai, & Fu, 2012). Taken together, these studies suggest that the interactions between parents and adolescents are nuanced and complex, and observational methodologies are well suited to answer questions regarding family processes.

Our findings do not provide an explanation of the mechanisms by which invalidation leads to self-injurious behaviors. As suggested by the biosocial theory, it is likely that parental invalidation impedes the development of critical emotion regulation skills and capacities. Of note, the unique association of parental invalidation and self-harm behaviors (and lack of association between parental invalidation and suicidal ideation) may also support the notion that self-harm may serve an important communication function. That is, given that one purpose of human emotions is to convey distress to others, if this communication is not adequately received, the individual may then need to escalate expressions of distress in order to be “heard” (Crowell et al., 2009).

The findings strengthen the empirical support that the parent-teen interactions affect future self-harm in vulnerable adolescents. Parental invalidation appears to result in increasing the intensity of teen’s expression of distress, which may explain the pattern of results holding for observation behavior but not significantly associated with suicidal ideation. Interestingly, high parental validation combined with high invalidation predicted the highest number of self-harm behaviors. These findings highlight the importance of using real interaction tasks to measure potentially complex transactions of behaviors between parents and teens versus self-reports, which may mask these nuances. It is also important to note that high levels of validation from parents does not compensate for high levels of invalidation, as these adolescents had the highest number of self-harm events. In fact, high levels of parental validation in combination with high levels of invalidation may be particularly distressing for teens, given that the parent is clearly capable of responding in appropriate ways, but does not do so consistently. This interplay between validation and invalidation is particularly important as it suggests that invalidation has a particularly pernicious effect on self-harm. The finding suggest that high levels of invalidation are not compensated by high levels of validation by parents, and perhaps any parent-teen interactions in the context of extreme invalidation may be distressing to the adolescent.

Given the theoretical importance of validation in the treatment of self-harm, the course of parental validation and invalidation over time was examined to determine the influence of time, treatment and their interaction during this clinical trial. Each treatment directly addressed parental validation according to its theoretical premise, however, neither yielded significant change, highlighting the need to develop additional strategies for change of this treatment target beyond the psychotherapy approaches studied in the RCT. Of interest, given that results of the RCT showed improvements in self-harm across conditions, with an additional advantage for DBT, it also suggests that positive treatment outcomes may not require changes in parental validation/invalidation as this was not the primary mechanism of change by which self-harm was reduced in this sample of adolescents. This highlights the importance of evaluating putative mechanisms of change, and the importance of continuing to work to improve the treatments that are disseminated for adolescents at risk for self-harm. We suspect that changing parent-adolescent interactions may be important for augmenting treatment response, and addressing the needs of those who do not respond to primarily individual therapy modalities. Additionally, parent validation may be a treatment target that required sustained focus, role-played practice, feedback, and significant clinical attention to change during treatment. Future research will need to be conducted to understand the most efficient methods of treatment to reduce self-harm; and identify the mechanisms of action. Future work is needed to determine if improvements in parental validation/invalidation may further augment positive treatment outcomes, especially given that treatment response rates were relatively low across conditions, with only 27% of youth in I/GST and-45% of youth in DBT ceased self-harm during treatment (McCauley et al., in press).

This study represents efforts to begin to tackle a modifiable parental behavior theoretically linked to the development of youth self-harm. The strengths to consider in this study include the longitudinal, carefully defined group of high risk adolescents, and observational coding of interactions which add to our confidence in our findings. However, it is important to note that the study had several limitations that affect its generalizability. The study had a small sample, combined suicidal and non-suicidal behaviors, relied upon a lab-based observation task for its assessment of parental behavior, and did not include a non-self-injuring comparison group. The small number of male participants precluded the evaluation of sex in analyses. Additionally, due to the low frequency of suicide attempts in the six month period, we collapsed non-suicidal self-injury and suicide attempts into one group and therefore could not distinguish unique relationships between parental responses and suicide attempts versus NSSI. The study was underpowered to detect group differences in treatment processes; however, post-hoc analyses indicated that the estimate of effect size on treatment and time was small, and 3138 participants would have been needed to reach significance. Consequently, we conclude that insufficient power due to a small sample was not the driving factor in findings supporting the null hypothesis for our second aim. Finally, given the study sample and design, the direction of effects is not yet determined. It is plausible that adolescent self-harm leads a parent to behave in invalidating ways or both parent and teen’s behaviors be caused by some other factor (e.g., other extreme family stress). Causality and direction of effects are important next steps in understanding how self-harm and parental behaviors may be optimally targeted and changed. Given these limitations, it is difficult to know if these findings generalize to other populations that did not share the characteristics of the sample. We believe the strengths of this study outweigh the limitations as this is the first study to our knowledge to employ an observational, longitudinal assessment of parent-child interaction patterns with adolescents at high risk for self-harm.

Research focused on family processes affecting self-harm is needed to drive prevention and intervention of these behaviors during adolescence. Our findings add to a small but growing literature that suggests that a problematic transaction that minimizes or dismisses emotional experiences fuels the occurrence and frequency of future self-harm in teens. This study assessed validation and invalidation over six months of treatment for adolescents at high risk for suicide, indicating that the interaction of parent validation and invalidation is significantly and robustly associated with self-harm. Importantly, this association was not significant when examining suicidal ideation and appears to thus be unique to self-harm behavior. These results highlight the importance of parental invalidation as a risk factor for self-harm behavior in adolescents. The evaluation of change in validation and invalidation during two treatment interventions specifically designed for high risk for suicide adolescents did not produce changes in observed validation patterns. It may be that for self-harm directly targeting emotion regulation skills, as opposed to the contextual factors contributing to emotional dysregulation, may be sufficient to improve adolescent functioning. It is likely that focused attention to altering parent’s responses to be more validating towards their teen’s emotional experience may require explicit targeting, in depth strategies, and sustained focus throughout the psychotherapy treatment process. Given the robust associations observed we would like to see future work develop brief and useful clinical strategies for parent-child interactions if it serves prioritized treatment targets and increases effective parent-teen interactions.

Table 3.

Mixed Effects Model Estimates for Time and Treatment on Validation and Invalidation

Parameter Validation Invalidation
Fixed Effects (β) β or σ
Estimate		 Standard
Error		 β or σ
Estimate		 Standard Error
Intercept 3.16** 0.4 2.82** 0.43
Treatment 0.14 0.59 0.32 0.61
Time 0.001 0.003 0.001 0.003
Time × Treatment 0.002 0.004 0.002 0.004
Random Effects (σ)
Within-person variability 0.4 0.63 0.66 0.81
Open in a new tab

Note.

**

p > .001

Public Significance Statement:

This study found a strong association between parent responses to their adolescents and subsequent self-harm. Additionally, it showed that two suicide-specific interventions were not effective in changing parental responses during treatment.

Acknowledgments

Disclosures/Acknowledgements: Work on this publication was supported American Foundation for Suicide Prevention, National Institute of Mental Health (NCT01528020; R01MH090159), Agency for Healthcare Research and Quality (AHRQ K12HS022982), awarded to the first author. Dr. Linehan has disclosures related to royalties received from DBT training materials. We would like to thank Alan Fruzzetti and Sheila Crowell for their assistance with the interaction task development and coding.

Biography

MOLLY ADRIAN received her PhD in developmental-clinical psychology from University of Maine. She is currently an assistant professor at the University of Washington in the Division of Child and Adolescent Psychiatry and Behavioral Medicine. Her areas of professional interest include emotion regulation, etiology and treatment of adolescent suicide risk behaviors.

MICHELE S. BERK received her Ph.D. in clinical psychology from New York University. She is currently Assistant Professor and Director of the Adolescent Dialectical Behavior Therapy Program in the Department of Psychiatry at the Stanford University School of Medicine. Her areas of professional interest include treatment approaches for adolescents at risk of suicide.

KATHRYN E KORSLUND received her PhD in clinical psychology from the Medical College of Pennsylvania at Hahnemann University. She is currently the Clinical Director of THIRA Health, LLC in Bellevue, WA. Her areas of professional interest focus on Dialectical Behavior Therapy (DBT) and include assessment of DBT adherence and DBT training, consultation and program development.

KATHRYN WHITLOCK received her MS in applied statistics from Purdue University. She is currently a biostatistician at the Seattle Children’s Research Institute. Her interests are in complex survey analysis methods and nonlinear modeling for quantitative outcomes.

ELIZABETH MCCAULEY received her PhD in developmental-clinical psychology from the State University of New York. She is currently a professor and the interim director of the Division of Child and Adolescent Psychiatry and Behavioral Medicine at the University of Washington and Seattle Children’s. Her areas of professional interest include differences in sex development and the prevention of adolescent depression and suicidality.

MARSHA LINEHAN received her PhD in psychology at Loyola University. She is currently a professor in the Department of Psychology and the Director of the Behavioral Research & Therapy Clinics at the University of Washington. Her areas of professional interest include Dialectical Behavior Therapy and treatment of suicidal behavior.

References

  1. Adrian M, & Berk M (in press). Behavioral Assessment of Emotion Dysregulation In, Crowell S & Beauchaine T (Editors) Oxford Handbook of Emotion Dysregulation. Oxford University Press. [Google Scholar]
  2. Adrian M, Zeman J, Erdley C, Lisa L & Sim L (2011). Emotional dysregulation and interpersonal difficulties as predictors of nonsuicidal self-injury in adolescent girls. Journal of Abnormal Child Psychology, 39, 389–400. [DOI] [PubMed] [Google Scholar]
  3. Asarnow JR, Porta G, Spirito A, Emslie G, Clarke G, Wagner KD, … Brent DA (2011). Suicide attempts and nonsuicidal self-injury in the treatment of resistant depression in adolescents: Findings from the TORDIA study. Journal of the American Academy of Child & Adolescent Psychiatry, 50(8), 772–781. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Bates D, Maechler M, & Bolker B (2013). lme4: Linear mixed-effects models using S4 classes. R package version 0.999999–0. 2012. URL: http://CRAN.R-Project.Org/package=lme4.
  5. Berk M, Adrian M, McCauley E, Asarnow J, Avina C, & Linehan M (2014). Conducting research on adolescent suicide attempters: Dilemmas and decisions. The Behavior Therapist / AABT, 37(3), 65–69. [PMC free article] [PubMed] [Google Scholar]
  6. Brain KL, Haines J, & Williams CL (1998). The psychophysiology of self-mutilation: Evidence of tension reduction. Archives of Suicide Research, 4(3), 227–242. [Google Scholar]
  7. Brent DA, Holder D, Kolko D, Birmaher B, Baugher M, Roth C, … Johnson BA (1997). A clinical psychotherapy trial for adolescent depression comparing cognitive, family, and supportive therapy. Archives of General Psychiatry, 54(9), 877–885. [DOI] [PubMed] [Google Scholar]
  8. Bridge JA, Goldstein TR, and Brent DA (2006). Adolescent suicide and suicidal behavior, Journal of Child Psychology and Psychiatry 47 (3), 372–394. [DOI] [PubMed] [Google Scholar]
  9. Centers for Disease Control and Prevention. (2016) Youth Risk Behavior Survey Data. Trends in Suicide-Related Behaviors. Available at: www.cdc.gov/yrbs. Accessed on August 4, 2017.
  10. Clarkin JF, Levy KN, Lenzenweger MF, & Kernberg OF (2007). Evaluating three treatments for borderline personality disorder: A multiwave study. American Journal of Psychiatry, 164(6), 922–928. [DOI] [PubMed] [Google Scholar]
  11. Cohen J, Cohen J, Cohen P, West SGA, Leona S, Patricia Cohen SGW, & Leona SA (2003). Applied multiple regression/correlation analysis for the behavioral sciences, Third Edition. Lawrence Erlbaum Associated; Mahwah, New Jersey. [Google Scholar]
  12. Cohen JA, Mannarino AP, & Knudsen K. (2005). Treating sexually abused children: 1 year follow-up of a randomized controlled trial. Child Abuse and Neglect, 29, 135–145. [DOI] [PubMed] [Google Scholar]
  13. Crowell SE, Baucom BR, McCauley E, Potapova NV, Fitelson M, Barth H, … Beauchaine TP (2013). Mechanisms of contextual risk for adolescent self-injury: Invalidation and conflict escalation in mother–child interactions. Journal of Clinical Child & Adolescent Psychology, 42(4), 467–480. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Crowell SE, Beauchaine TP, & Linehan MM (2009). A biosocial developmental model of borderline personality: Elaborating and extending linehan’s theory. Psychological Bulletin, 135(3), 495–510. [DOI] [PMC free article] [PubMed] [Google Scholar]
  15. Crowell SE, Beauchaine TP, McCauley E, Smith CJ, Vasilev CA, & Stevens AL (2008). Parent-child interactions, peripheral serotonin, and self-inflicted injury in adolescents. Journal of Consulting and Clinical Psychology, 76(1), 15–21. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Curtin S, Warner M, & Hedegaard H (2016). Increase in Suicide in the United States, 1999–2014. NCHS Data Brief, (241), 1–8. [PubMed] [Google Scholar]
  17. Dawson JF (2014). Moderation in management research: What, why, when, and how. Journal of Business and Psychology, 29(1), 1–19. [Google Scholar]
  18. Fixsen DL, Naoom SF, Blase KA, Friedman RM & Wallace F (2005). Implementation Research: A Synthesis of the Literature. Tampa, FL: University of South Florida, Louis de la Parte Florida Mental Health Institute, The National Implementation Research Network. [Google Scholar]
  19. Fruzzetti AE (2010). Validating and invalidating behavior coding scale manual. Report, University of Nevada, Reno. [Google Scholar]
  20. Garber J, Little S, Hilsman R, & Weaver KR (1998). Family predictors of suicidal symptoms in young adolescents. Journal of Adolescence, 21(4), 445–457. [DOI] [PubMed] [Google Scholar]
  21. Glenn CR, Franklin JC, & Nock MK (2015). Evidence-based psychosocial treatments for self-injurious thoughts and behaviors in youth. Journal of Clinical Child & Adolescent Psychology, 44(1), 1–29. [DOI] [PMC free article] [PubMed] [Google Scholar]
  22. Gratz KL, Conrad SD, & Roemer L (2002). Risk factors for deliberate self-harm among college students. American Journal of Orthopsychiatry, 72(1), 128–140. [DOI] [PubMed] [Google Scholar]
  23. Hawton K, Saunders K, O’Connor R (2012). Self-harm and suicide in adolescents. The Lancet, 379, 23–29. [DOI] [PubMed] [Google Scholar]
  24. Johnson JG, Cohen P, Gould MS, Kasen S, Brown J, & Brook JS (2002). Childhood adversities, interpersonal difficulties, and risk for suicide attempts during late adolescence and early adulthood. Archives of General Psychiatry, 59(8), 741–749. [DOI] [PubMed] [Google Scholar]
  25. Kazdin AE (2007). Mediators and mechanisms of change in psychotherapy research. Annual Review of Clinical Psychology, 3(1), 1–27. [DOI] [PubMed] [Google Scholar]
  26. Klonsky ED, May AM, & Glenn CR (2013). The relationship between nonsuicidal selfinjury and attempted suicide: converging evidence from four samples. Journal of Abnormal Psychology, 122(1), 231. [DOI] [PubMed] [Google Scholar]
  27. Linehan M (1993). Cognitive-behavioral treatment of borderline personality disorder. Guilford Press. [Google Scholar]
  28. Linehan MM, Comtois KA, Brown MZ, Heard HL, & Wagner A (2006). Suicide Attempt Self-Injury Interview (SASII): Development, reliability, and validity of a scale to assess suicide attempts and intentional self-injury. Psychological Assessment, 18(3), 303–312. [DOI] [PubMed] [Google Scholar]
  29. Linehan MM, Dimeff LA, Reynolds SK, Comtois KA, Welch SS, Heagerty P, & Kivlahan DR (2002). Dialectical behavior therapy versus comprehensive validation therapy plus 12-step for the treatment of opioid dependent women meeting criteria for borderline personality disorder. Drug and Alcohol Dependence, 67(1), 13–26. [DOI] [PubMed] [Google Scholar]
  30. Linehan M, McCauley E, Berk M, Asarnow J, Korslund K, Adrian M, … Botanov Y (2016). Dialectical Behavior Therapy compared to supportive treatment: A randomized controlled trial for adolescents at high risk for suicide. Symposium presented at the New Outcome Data on Treatments for Suicidal Adolescents Associations for Behavioral and Cognitive Therapies, New York, NY. [Google Scholar]
  31. Linehan MM, Heard HL, & Armstrong HE. (1993). NAturalistic follow-up of a behavioral treatment for chronically parasuicidal borderline patients. Archives of General Psychiatry, 50(12), 971–974. [DOI] [PubMed] [Google Scholar]
  32. Linehan MM, Wagner AW, Tutek D. (1990) DBT Expert Rating Scale. University of Washington; Seattle, WA. [Google Scholar]
  33. McCauley E, Berk M, Asarnow J, Adrian M, Korslund K, Cohen J,…Linehan MM (in press). Efficacy of Dialectical Behavior Therapy for adolescents with high risk for suicide: A randomized clinical trial. [DOI] [PMC free article] [PubMed]
  34. McMain SF, Links PS, Gnam WH, Guimond T, Cardish RJ, Korman L, & Streiner DL (2009). A randomized trial of dialectical behavior therapy versus general psychiatric management for borderline personality disorder. American Journal of Psychiatry, 166(12), 1365–1374. [DOI] [PubMed] [Google Scholar]
  35. Mehlum L, Ramberg M, Tørmoen AJ, Haga E, Diep LM, Stanley BH, … Grøholt B (2014). Dialectical behavior therapy compared with enhanced usual care for adolescents with repeated suicidal and self-harming behavior: Outcomes over a one-year follow-up. Journal of the American Academy of Child & Adolescent Psychiatry, 55(4), 295–300. [DOI] [PubMed] [Google Scholar]
  36. Miller AL, Rathus JH, & Linehan MM (2006). Dialectical behavior therapy with suicidal adolescents. Guilford Press. [Google Scholar]
  37. Mościcki EK (2001). Epidemiology of completed and attempted suicide: toward a framework for prevention. Clinical Neuroscience Research, 1(5), 310–323. [Google Scholar]
  38. Nock MK, & Mendes WB (2008). Physiological arousal, distress tolerance, and social problem-solving deficits among adolescent self-injurers. Journal of Consulting and Clinical Psychology, 76(1), 28–38. [DOI] [PubMed] [Google Scholar]
  39. Core Team R. (2008). R: A language and environment for statistical computing. R foundation for statistical computing. Vienna, Austria.
  40. Raudenbush SW, & Bryk AS (2002). Hierarchical linear models: Applications and data analysis methods. SAGE. [Google Scholar]
  41. Reynolds WM (1988). Suicidal ideation questionnaire: Professional Manual. Psychological Assessment Resources. [Google Scholar]
  42. Reynolds WM, & Mazza JJ (1999). Assessment of suicidal ideation in inner-city children and young adolescents: Reliability and validity of the Suicidal Ideation Questionnaire-JR. School Psychology Review, 28(1), 17. [Google Scholar]
  43. Ross S, & Heath N (2002). A study of the frequency of self-mutilation in a community sample of adolescents. Journal of Youth and Adolescence, 31(1), 67–77. [Google Scholar]
  44. Saarni C (1999). The development of emotional competence Guilford Press; New York, NY. [Google Scholar]
  45. Shenk CE, & Fruzzetti AE (2011). The Impact of Validating and Invalidating Responses on Emotional Reactivity. Journal of Social and Clinical Psychology, 30(2), 163–183. [Google Scholar]
  46. Skegg K (2005). Self-harm. The Lancet, 366(9495), 1471–1483. [DOI] [PubMed] [Google Scholar]
  47. Sterne JA, White IR, Carlin JB, Spratt M, Royston P, Kenward MG, … Carpenter JR (2009). Multiple imputation for missing data in epidemiological and clinical research: potential and pitfalls. Bmj, 338, b2393. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Strodtbeck FL (1951). Husband-wife interaction over revealed differences. American Sociological Review, 16(4), 468–473. [Google Scholar]
  49. Suyemoto KL (1998). The functions of self-mutilation. Clinical Psychology Review, 18(5), 531–554. h [DOI] [PubMed] [Google Scholar]
  50. van den Bosch LM, Verheul R, Schippers GM, & van den Brink W (2002). Dialectical behavior therapy of borderline parents with and without substance use problems. Implementation and long term effects. Addictive Behavior, 27(6), 911–923. [DOI] [PubMed] [Google Scholar]
  51. Wedig MM, & Nock MK (2007). Parental Expressed Emotion and Adolescent Self-Injury. Journal of the American Academy of Child & Adolescent Psychiatry, 46(9), 1171–1178. [DOI] [PubMed] [Google Scholar]
  52. You J, Leung F, Lai CM, & Fu K (2012). The associations between non-suicidal selfinjury and borderline personality disorder features among chinese adolescents. Journal of Personality Disorders, 26(2), 226–237. [DOI] [PubMed] [Google Scholar]

RESOURCES