Using media to regulate our moods: When we turn to television and music

I came across an interesting study today in a recent issue of Personality and Individual Differences. Dara Greenwood and Christopher Long (2009) examined which individuals are most likely to utilize particular types of media (e.g., television, music) in response to specific emotional states. Different individuals are impacted differently by their emotions. As such, insight into which individuals are likely to utilize particular methods for regulating their emotions would be useful in treatment settings, as it would help clinicians to tailor skills training to the client while also revealing potential vulnerabilities to harmful behaviors.
In this particular study, the authors had a sample of 229 undergraduates fill out questionnaires detailing their ability to regulate emotions [Difficulties in Emotion Regulation Scale (DERS); Gratz & Roemer, 2004]. The authors also had participants respond to a series of questions regarding the use of entertainment media within the context of emotional states, including positive emotions, negative emotions, and boredom. Extending this further, the authors asked each participant how often they utilized particular types of media (music, television, movies, magazines, and video games) in each of these states.
The authors found that, in both positive and negative mood states, music was the most frequently used form of entertainment media. In negative moods, television was the second most common form of media use, whereas in positive moods, television and movies were the second most frequently used forms of media. In states of boredom, television was the most frequently used form of entertainment media.
Perhaps more interestingly, the authors found that individuals who experience difficulty regulating their emotions and who frequently ruminate were the most likely to watch television when upset. The authors argue that individuals choose this method of addressing their emotions because it allows for passive participation in an activity that might distract from aversive self-awareness, which is consistent with other theories regarding the use of particular behaviors in response to negative affective states (e.g., Heatherton & Baumeister, 1991). Another more complex finding was that both individuals with low emotion regulation abilities and high emotion regulation abilities were likely to turn to music when upset. The authors use a more nuanced argument in explaining this finding - one that I believe is extremely compelling. Individuals high in emotion regulation skills, they argue, use music to help them reflect upon their mood and better understand it in a healthy manner. Individuals low in emotion regulation skills, on the other hand, use music in an attempt to simply distract from what they are feeling, often resulting in further brooding and deteriorating mood.
In my opinion, this study is a useful foundation upon which further work must be done in order for us to understand these phenomena to a sufficient degree. In this sense, it is a very useful study that provides some interesting findings, but it leaves several important questions unanswered. Primarily, I found myself wondering the following:

1. To what degree is entertainment media the first thing that these individuals turn to when upset? Might they use alcohol, drugs, binge eating, non-suicidal self-injury, or another behavior first?
2. To what degree does an undergraduates sample with no measurement of psychopathology provide an adequate understanding of how these behavioral tendencies appear outside the lab?
3. To what degree do these behaviors actually work? In other words, do these individuals turn to entertainment media for a brief time, give up on it as a tool, then engage in another behavior or do they actually experience significant alterations in their mood?
4. What types of choices are these individuals making within each type of media? Are they listening to sad songs when sad or listening to happier music in attempt to lift their spirits?
5. What type of cognitive approach are these individuals taking when engaging in these activities? Are they being mindful of the media, focusing their attention only on what they see or hear and thus allowing their emotions to run their course or are they ruminating as they watch or listen, extending negative moods and increasing their vulnerabilities to more destructive behaviors?

I ask these questions not as a criticism of the study - no study ever answers all questions and, in fact, good ones prompt readers to reflect upon the findings and pursue follow-up data. I ask them, rather, because the study serves as an excellent corollary to our ongoing series of articles on dialectical behavior therapy. Part 3, which will appear later this week, will deal with emotion regulation skills, which can be defined as the manner in which individuals attempt to adjust or sustain their current emotional states (Thompson, 1994). Studies like this provide a framework within which to consider how individuals approach their emotions, how motivated they are to change them, and to what degree they show a tendency to use methods for emotion regulation that actually backfire.
In this sense, my fourth and fifth questions are particularly important to consider. If I am sad and decide to listen to "Everybody Hurts," by REM, that is likely to have a decidedly different impact on my emotions than will listening to "Walking on Sunshine" by Katrina and the Waves. If individuals who have difficulty regulating their emotions choose to listen to upsetting songs when upset, the question of which type of entertainment media used (e.g., music versus television) becomes less important than questions regarding the content of media used. In clinical experience, I have found that, quite often, clients need to be reminded that an episode of "Friends" would be a better tool for distraction and emotion regulation than will an episode of "Six Feet Under," even though the latter is a beautifully written program. The argument is not that sad media is a bad thing, but simply that sad media, when viewed or heard while sad, may not be the most effective tool for that particular moment. Of course, empirical data demonstrating this effect would be more useful than my clinical lore.
With respect to the fifth question - what type of cognitive approaches are individuals taking when they utilize entertainment media in response to emotions - the answer would require a much more complex experimental procedure. If individuals are able to control their attention by focusing on music, television, or any other form of media, utilizing all of their cognitive resources on positive or neutral stimuli until their own emotions run their course (or at least decrease in intensity), this is a very healthy behavioral choice, as it slows down impulsive responses and better positions individuals to effectively problem solve. If, on the other hand, entertainment media induces or prolongs rumination, the behavior will have the opposite effect.
In Part 3 of the DBT series, I will explain a variety of emotion regulation skills suggested by Linehan. In the meantime, I thought it would be useful and interesting to consider how we use entertainment media in the context of our emotions and to what degree our choices in that regard are healthy ones. Remember, while DBT is used for the treatment of borderline personality disorder, bulimia, and other mental illnesses, we all experience emotions and must make decisions regarding how to alter or sustain them. Studies like this might provide us with an added perspective, allowing for greater insight into why, on occasion, we are more or less successful at effectively changing how we feel.

Pulling hair to feel better: Emotion regulation in trichotillomania

There is simply no way to cover every DSM diagnosis in great detail regularly on PBB, but we like to make a legitimate effort to touch upon as many as possible. One disorder yet to be subject to any attention on thus far on the site is trichotillomania (TTM). TTM is characterized by the repetitive pulling of one's own hair and is estimated to impact approximately 1-2.5% of the population (Christianson, Pyle, & Mitchell, 1991). Depending upon the strictness with which the criteria for TTM are defined, that estimate can balloon to as high as 15% (Rothbaum, Shaw, Morris, & Ninan, 1993). Hair pulling is not limited to the hair on one's head and some individuals pull hair from multiple locations. The resulting impact on physical appearance in severe cases can be highly upsetting for the individual struggling with TTM and, as such, the disorder can have a highly problematic impact on an individual's social life.

In a study just released in Behaviour Research and Therapy, Anna Shusterman, Lauren Feld, Lee Baer, and Nancy Keuthen (2009) utilized data from a massive online survey to examine the role that emotions play in prompting and sustaining this disorder. The description of TTM in the DSM-IV-TR as well as a number of prior studies have linked TTM behaviors with a sense of relief on the part of the individual exhibiting the behavior. In other words, many individuals have reported that pulling out their own hair has resulted in immediate decreases in negative emotions. Despite this potentially valuable function, the behavior also includes a variety of less comfortable correlates, namely subsequent feelings of shame and guilt and a strong desire to cease the behavior. So, the behavior becomes rewarding through its ability to quickly reduce certain negative emotions, but also results in several emotional and social consequences.

If you have read any of our articles on binge eating, you might notice a parallel here - individuals who engage in dysregulated behaviors when upset because the behavior offers immediate relief from aversive emotions, but who then subsequently regret the behavior and wish they could escape the pattern. Keep this parallel in mind as you read the rest of this article, as it is an important consideration that also generalizes to several other dysregulated behaviors (e.g., non-suicidal self-injury, drinking alcohol to cope with negative affect).

Shusterman and colleagues (2009) used the terms "pullers" and "non-pullers" throughout the paper to refer to individuals who do and do not pull their own hair and were interested in testing three hypotheses. First, they believed that pullers would endorse greater difficulty in regulating emotions than would non-pullers. Second, they believed that the degree to which pullers reported difficulty regulating emotions would correspond to the degree to which they pull their own hair, with greater difficulty regulating emotions resulting in greater levels of hair pulling. Third, they believed that the degree to which a puller reports difficulty regulating a specific emotion would correspond with the degree to which that particular emotion serves as a prompt for hair pulling. In other words, if a puller indicated that she generally struggles to regulate anger, the authors anticipated that anger would be a common cue that led to her pulling her hair.

The data for this study was accumulated online through a massive survey project and participants were recruited through advertisements and word of mouth. In total, the sample included 1,162 "pullers" and 175 "non-pullers." All of the data reflected participants' responses to self-report questionnaires. Additionally, no formal diagnostic procedures were utilized, as the authors believed it would be more useful to examine trichotillomania as a continuum rather than simply looking at folks who do or do not meet criteria for the disorder based upon the DSM-IV-TR. Despite these limitations, we believe the study offered an interesting glimpse at the role of emotions in this particular behavior.

Consistent with their hypotheses, the authors found that pullers reported greater difficulty "snapping out" of emotions than did non-pullers. "Snapping out" was not clearly defined, which raises legitimate questions regarding the validity of their assessment, but nonetheless, this represents preliminary evidence that individuals exhibiting symptoms of trichotillomania have a particularly difficult time effectively altering emotional states in a healthy manner. Additionally, the authors found that the degree to which pullers reported difficulty regulating emotions correlated with the severity of their hair pulling behavior. So, pullers who had a particularly difficult time regulating emotions typically exhibited more severe hair pulling behavior. Additionally, this relationship held even when comparing the individuals with the mildest level of hair pulling to non-pullers, with mild pullers reporting significantly greater difficulty regulating emotions.

Perhaps the most interesting finding, however, was related to the authors' third hypothesis. In this sample, they found that individuals were often prompted to pull their hair in response to the specific emotions they reported having the most difficulty regulating. In other words, if an individual who pulls his hair struggles to regulate feelings of boredom, than boredom is likely to be a particularly salient cue that prompts him to pull his hair.

Importantly, as I hinted at above, the online self-report measures represent a weakness in examining these particular questions. Asking somebody how they typically feel right before and right after a behavior that occurred in the past is an unreliable form of measurement. We simply are not particularly accurate when we try to recall this type of information over extended periods of time. A better approach would involve taking real time measurements of mood, etc...over a period of time and then using more sophisticated statistical analyses such as hierarchical linear modeling to examine the data. As such, this study should be seen as preliminary and its results should be interpreted with caution; however, there is nothing wrong with preliminary data as long as folks do not overstep the bounds of what such data care capable of telling us and the authors did nothing of the sort here.

As I read this article, I could not help but reflect upon the similarities between hair-pulling and a variety of other problematic behaviors. In past PBB articles, we have covered these behaviors on their own, but we have also called attention to variables like negative urgency and distress tolerance that have been shown to predict a variety of problematic behavioral outcomes. Thus far, to our knowledge, there is no empirical data linking distress tolerance or negative urgency to hair pulling, but it seems likely that individuals who pull their hair in an effort to regulate emotions likely exhibit elevations in these same risk factors. If so, this has important treatment implications, as clinicians could stress emotion regulation skills and test the utility of emotion-based therapeutic approaches such as dialectical behavior therapy (DBT) and acceptance and commitment therapy (ACT). In fact, as Shusterman and colleagues pointed out in their article (2009), there is preliminary evidence indicating that ACT might be an effective treatment capable of reducing hair pulling behaviors (Begotka, Woods, & Wetterneck, 2004; Woods & Twohig, 2008).

A question left unanswered by this and similar studies on other problematic behaviors is why emotion regulation difficulties would lead to this particular behavior. As a researcher focused on treatment, I often de-emphasize this particular question. I do this for several reasons. First, treatments such as DBT appear effective for reducing disparate behaviors (e.g., NSSI, binge eating) that are prompted by difficulties regulating emotions. As such, the reason an individual chose a particular behavior does not appear to impact the effectiveness of treatment, at least in certain circumstances. Second, I believe that the "insight fallacy" is one of the primary weaknesses of most non-scientific approaches to psychotherapy. The belief that understanding the origins of a behavior (beyond risk factors such as negative urgency or other prompting events) is a necessary and sufficient means for altering that behavior has simply not stood up to testing and, as such, while there is nothing harmful about such insights, I often do not emphasize them in my attempts to consider treatments. Nonetheless, it remains interesting to consider why one individual may resort to NSSI, another might binge eat, and another might pull hair in efforts to regulate emotions. For some, I suspect the initial episodes result from peer or family modeling and the inherently reinforcing emotion regulation properties of the behavior then sustain it for those who are vulnerable. For others, I think might simply be chance. One day, they are upset and they happen to stumble upon a behavior that works, drastically increasing the chances that the behavior will be repeated in the future. I'm curious what you think about this question. Is there a specific reason why some individuals choose one behavior while others choose another?

Struggling with feelings: A look at emotion dysregulation in borderline personality disorder

None of us enjoy feeling negative emotions. By definition, it is an aversive experience, and given the option, all of us would generally prefer to avoid feeling distress. That being said, most people are able to manage things when they are upset, navigating through uncomfortable sensations to accomplish goals and staying with their negative emotions as needed in order to avoid costly consequences. For others, however, negative emotions serve as an almost insurmountable obstacle, or at least seem that way while the experience is actively occurring.

The impact of emotions on behavioral outcomes and suicide risk is the focus of my research, so I am always excited to read about new studies in this area. Today, I want to discuss a study just published in a special issue of Personality Disorders: Theory Research and Practice by Kim Gratz, Zachary Rosenthal, Matthew Tull, Carl Lejuez, and John Gunderson (2009) looking at emotion dysregulation in borderline personality disorder (BPD). You will recall from prior PBB articles on BPD, individuals who struggle with this diagnosis are often characterized by rapidly shifting emotions and a reduced tolerance for negative emotions that leads to the frequent use of harmful behaviors (e.g., self-injury) in an attempt to reduce such feelings. In a past article, we described evidence for this tendency through a discussion of a publication by Nock and Mendes (2008) and I hope you will consider reading that article as a supplement to today's discussion.

In the current study, Gratz et al. (2009) wanted to experimentally measure the degree to which individuals with BPD actually exhibit difficulties regulating emotions relative to individuals without BPD. Even though low distress tolerance and a tendency to engage in harmful behaviors in response to negative emotions are central facets of our theoretical understanding of BPD (Linehan, 1993), few studies have directly tested this assumption and the bulk of those that have done so have utilized self-report measures of emotion dysregulation rather than directly observing the phenomenon (e.g., Livesley et al., 1998).

For the purpose of this study, the authors defined emotion dysregulation as a combination of the following four characteristics (see page 19 of the actual article or consult Gratz & Roemer, 2004):

• "Lack of awareness, understanding, and acceptance of emotions."

• "Lack of access to adaptive strategies for modulating the intensity and/or duration of emotional responses."

• "An unwillingness to experience emotional distress as part of pursuing desired goals."

• "The inability to engage in goal-directed behaviors when experiencing distress."

Gratz and colleagues (2009) wanted to directly test whether individuals with BPD struggled particularly with the final two of those four components and designed a study that included 35 participants, 17 with BPD and 18 without BPD. The focal point of the study was a computer task known as the Paced Auditory Serial Addition Task - Computerized (PASAT-C; Lejuez, Kahler, & Brown, 2003), which requires participants to perform basic math computations at an increasingly fast pace. Each time a participant makes an error, a loud and aversive explosion noise is heard. Toward the end of the task, the pace of the problems is quick enough to be essentially impossible, thereby assuring that the participant is bombarded with aversive noises signally poor performance on the task. Eventually, the participant is given the option to quick the task and the length of time that the participant continues is used as a measure of their willingness to persist in goal directed behavior while distressed (actual measures of distress are also included to ensure that quitting does not simply signify boredom). In the case of this particular study, participants were told that the longer they persisted, the more time they would have on a subsequent task and that their performance on that later task would determine how much money they earned for participating in the study. This was actually a bit of deception, as all participants were given equal compensation, but the researchers wanted to ensure that participants were motivated to continue in the task.

Before analyzing their data, Gratz and colleagues (2009) hypothesized that individuals with BPD would be more likely to quit the task early, an indication that they are less willing to experience distress for the sake of a goal, and that they would perform worse on the final round of the PASAT-C, an indication that they struggle to engage in goal-directed behaviors when distressed. Their results supported some, but not all of their hypotheses.

Despite the fact that both groups reported equal levels of distress in response to the PASAT-C, individuals with BPD were more likely to quit the task early than were individuals without BPD. In other words, it is not that individuals with BPD necessarily feel more distress, but rather that the experience of distress is viewed as less tolerable. So, while most individuals are willing to experience moderate to severe levels of distress when a task seems important enough to warrant discomfort, individuals with BPD have a harder time drawing that same conclusion and instead prioritize reducing negative emotions by whatever means necessary, even if there are both short and long-term costs associated with their efforts to escape.

Contrary to hypotheses, individuals with BPD did not perform worse on the task. As such, the evidence did not support the notion that individuals with BPD struggle to succeed in tasks when they do persist through the experience of negative emotions. In all honesty, this did not surprise me, although I do understand why the authors (who know significantly more about that topic than I do) believed it might be true. BPD is not characterized by failures in performances, but rather difficulties managing the experience of negative emotions. In this sense, I think that individuals with BPD are more likely to be characterized by the belief that they will falter or currently are faltering when upset than they are to actually struggle with a task. In this sense, they are plagued by distorted automatic negative thoughts in need of remediation.

Gratz and her colleagues (2009) were careful to explain several limitations in their study, some of which warrant mentioning here. First of all, the sample size was quite small and, additionally, the excluded individuals who met criteria for depression. Given that a substantial number of individuals with BPD are also depressed (Skodol et al ., 2002) and depression could impact task performance, it is possible that this significantly impacted the results. Additionally, although the PASAT-C is a fascinating task that has provided some important findings, it is unlikely to parallel the types of experiences that prompt strong emotional responses from individuals with BPD outside the lab and, as such, might not be the best measure for examining emotional responses in BPD. In the Nock and Mendes (2008) study that we described in a prior article, the distress was primarily interpersonal in nature and the findings were a bit more consistent with those hypothesized here. As someone who has used the PASAT-C in laboratory tasks before, I am aware of how distressing it is for the participant, but I do wonder at times how it compares to more interpersonally-based tasks like the Distress Tolerance Test.

Despite these limitations, this study had obvious value. First of all, the authors made a strong effort to experimentally measure a topic - emotion dysregulation - that often only gets examined through self-report, a useful but inferior form of measurement. Secondly, the authors administered structured clinical interviews to obtain their diagnoses, an approach we have described before as the most valid and reliable way to reach a diagnostic decision. Studies like this help us make sure that our assumptions about particular diagnoses actually reflect reality and we can use data like these to further our understanding of mental illness and to help designed treatments that target the most impairing symptoms of a given disorder.

Regulating emotion after experiencing a sexual assault

Philadelphia, PA, 22 October 2009 - After exposure to extreme life stresses, what distinguishes the individuals who do and do not develop posttraumatic stress disorder (PTSD)? A new study, published in the October 1st issue of Biological Psychiatry, suggests that it has something to do with the way that we control the activity of the prefrontal cortex, a brain region thought to orchestrate our thoughts and actions.

Researchers at the Mount Sinai School of Medicine examined women who had been the victims of violent sexual assault, some of whom developed PTSD and others who did not develop any serious emotional symptoms afterwards. Using a brain imaging technique, they evaluated the ability of these women to voluntarily modify their own responses to unpleasant emotional stimuli and found that it was the trauma history itself, not how well they endured this sort of trauma, that influenced their ability to dampen subsequent emotional responses.

Surprisingly, however, the ability of the subjects to amplify their emotional responses to unpleasant stimuli was related to psychological outcome after the sexual assault. The resilient individuals, that is, those who endured sexual assault without developing emotional symptoms, were able to enhance the activation of emotional brain circuitry in response to unpleasant stimuli more than either those with PTSD or healthy controls who had never experienced a serious sexual assault.

Corresponding author Dr. Antonia New explained the findings: "This raises the possibility that the ability to focus on negative emotions permits the engagement of cognitive strategies for extinguishing negative emotional responses, and that this ability might be related to resilience. This is important, since it has implications for how we might enhance resilience."

These findings suggest that exposure to extremely stressful situations may leave an "emotional scar" that may influence the capacity to be resilient to the impact of subsequent stressors, even when one does not develop PTSD. "These data seem to support an idea that has emerged from clinical descriptions of resilient people, i.e., that people who are resilient are able to be flexible in the way that they respond to changing emotional contexts. It would be helpful to know how we can enhance the flexible activation of these prefrontal cortex networks in people with compromised resilience," commented Dr. John Krystal, Editor of Biological Psychiatry.

Dr. New agrees, adding that "perhaps the enrichment of the broad capacity to tolerate negative emotional experiences might be helpful in promoting resilience. Further work needs to be done on whether the feature of this capacity that relates to resilience is about the ability to tolerate one's one responses, or whether it is the ability to respond distress in others."